Wikipedia

Wikipedia:Village pump (technical)/Archive 82

< Wikipedia:Village pump (technical)
Village pump (technical) archive  

This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.

< Older discussions · Archives: A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P, Q, R, S, T, U, V, W, X, Y, Z, AA, AB, AC, AD, AE, AF, AG, AH, AI, AJ, AK, AL, AM, AN, AO, AP, AQ, AR, AS, AT, AU, AV, AW, AX · 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212

Bots and spam blacklist

Anyone know if there is a way to allow a bot to edit past blacklist restrictions? I finally figured out why CSBot sometimes goes crazy: once every so often it tries to point out a copy where the source is on the blacklist. The edit fails, so return to step one.

I can just ignore tagging when it hits a blacklisted site, but those are arguably even more important to tag. — Coren (talk) 14:42, 6 November 2010 (UTC)

The blacklist also affects Twinkle users, who because of limitations of the script, do not get an error message. Of course, the logical workaround is to check if the edit was successful, and if not, omit the offending URL. PleaseStand (talk) 17:32, 6 November 2010 (UTC)
You could also wrap the problem URL in <nowiki></nowiki> tags, as done here for example. Anomie⚔ 05:18, 7 November 2010 (UTC)

overlapping lines

Template:Opinion paints the background around the text to whatever color you choose for the background but

if you set your line height to 1 then

it has the surprising effect of painting right over the dangling parts (gjpqy) of the text on the line above.

i.e. the top part of one line overlaps the bottom part of the line above.

If you put the text into a table and set the background to some color then it doesnt do this.

Do any of you guys see any way of modifying fontcolor so that it doesnt do this

(or some way of setting somekind of line height attribute in my css to prevent lines from overlapping?)

Lemmiwinks2 (talk) 01:29, 6 November 2010 (UTC)

I think you are referring to descenders. The issue is probably with {{fix}}. Example please. ---— Gadget850 (Ed) talk 18:25, 6 November 2010 (UTC)

http://i205.photobucket.com/albums/bb102/nooneparticular/line-height.png

-{{fontcolor||red|┼┼┼┼┼┼┼┼Bp}} ┼┼┼{{fontcolor||yellow|┼┼┼┼┼┼┼┼Bp}} -{{fontcolor||red|┼┼┼┼┼Bp}}

the character is a box drawing element so it should be the full height and width of the character. Lemmiwinks2 (talk) 07:12, 7 November 2010 (UTC)

that picture doesnt really show it very well.

when I use 36px font with 5px top and bottom margins it works perfectly.

when I use small fonts (and I like to use very small fonts) it gets screwy.

all I can do it fiddle with the margins and see if I can get it to work reasonably well.

Lemmiwinks2 (talk) 08:50, 7 November 2010 (UTC)

External link comes back to Wikipedia article

On the article Eoghan Quigg there is an external link, www.eoghanquiggmusic.com which takes you back to the Wikipedia article. Anyone any idea what's going on? DuncanHill (talk) 19:51, 6 November 2010 (UTC)

Looks like some one registered the domain and redirected it to the article. We can't do anything about the site, but the link is circular. ---— Gadget850 (Ed) talk 20:03, 6 November 2010 (UTC)
The source code of the website sets it all up. I removed the link from the infobox and from the external links.--Bbb23 (talk) 20:12, 6 November 2010 (UTC)
It appears from Google's cache, http://www.eoghanquiggmusic.com/tesco/ and links from other sites that it really was the official site for Eoghan Quigg before most of the domain was redirected to Wikipedia. Perhaps by the subject due to career failure? PrimeHunter (talk) 20:50, 6 November 2010 (UTC)
OK, thanks. DuncanHill (talk) 12:34, 7 November 2010 (UTC)

Vector bug

Some time ago I saw that, in the vector skin, it was sometimes loading the page as I was logged-out. Now I've seen that bug appearing again. What is the problem? Is it a cache issue? It is really scary when I get fooled by the fact that the page is loaded as I'm logged-out, when in reality I'm still logged in (if I refresh the page, I am logged in). HeyMid (contributions) 22:15, 6 November 2010 (UTC)

Sometimes the session data are lost. BTW, I use Monobook and it happened to me so it is not a Vector bug. – Allen for IPv6 23:37, 6 November 2010 (UTC)
Is there a way to fix this? Also, can this be compared with the session data lost message which sometimes appears when saving a page? HeyMid (contributions) 09:16, 7 November 2010 (UTC)

Text becomes "stretched" when non-English characters are present

When certain non-english characters are present, the text of paragraph that they are included in seems to stretch. For instance, look at the second paragraph on this section of the article on Prince's Lovesexy record. Notice everything from the start ("Side two...") until the beginning of italics (Camille-like vocals) looks different from the rest of the content of the page? Notice the inclusion of a special peace-sign symbol which makes up the title of another album. I've seen it with other characters (often Japanese or Chinese, sometimes other non-Arabic character languages) on other articles (the one above is one example of many). Anyone else notice it? Doc StrangeMailboxLogbook 06:18, 7 November 2010 (UTC)

Such an issue is with your browser, operating system and/or fonts installed on your system. —TheDJ (talk • contribs) 12:38, 7 November 2010 (UTC)

my font size decreased

i was looking at an article and i tried to type ctrl F to find something but i hit something else and made the font size smaller. how do you get it back?--Gaius Claudius Nero (talk) 16:34, 7 November 2010 (UTC)

In Firefox, at least, try pressing control-0 (that's control-zero, not control-O). I think that shortcut works with many other browsers too, although I'm not sure about all. You can use control-mousewheel to change the size in general. --ais523 16:45, 7 November 2010 (UTC)
On my version (Firefox 3.0) it's Ctrl-minus to shrink and Ctrl-plus to expand. Antandrus (talk) 16:47, 7 November 2010 (UTC)
okay, it worked. i see how it happened. thanks.--Gaius Claudius Nero (talk) 16:47, 7 November 2010 (UTC)

New Census Figures Jan 2011

Question about Wikipedia planning for census data on all United States pages. The demographics will all be out of date as of January 2011, when the U.S. Census Bureau releases the 2010 census figures. That's a lot of instantaneously outdated information on Wikipedia. Does Wikipedia have plans to run an update Bot? I have noticed two different existing methods on the pages, one possibly due to human error. Taking Scotts Bluff, Nebraska as an example.

  • A lot of those pages correctly reference the U.S. Census Bureau Fact Sheet for their demographics.
  • Others, such as Scotts Bluff, Nebraska, don't reference the correct URL address to pick up the figures for the zip code demographics, so when clicking the reference, you only get the American Fact Finder main page.

Any insight out there as to what happens January 2011? --Maile66 (talk) 18:58, 7 November 2010 (UTC)

Please re-post at the talk page of WP:2010 US Census. — Dispenser 21:04, 7 November 2010 (UTC)
Done. Thanks for pointing me to the right place. Maile66 (talk) 21:46, 7 November 2010 (UTC)
You might contact User:Ram-Man to see if he's planning on updating his bot, which did the initial census parsing. Though he hasn't posted since June, so you might want to send him an email. Corvus cornixtalk 22:36, 7 November 2010 (UTC)
Done. It's worth a try. Thanks for the suggestion.
--Maile66 (talk) 23:04, 7 November 2010 (UTC)

Number of page revisions of Wikipedia:Sandbox

I have been trying to find the number of revisions made to the Sandbox. I have tried using X!'s tool but there are to many revisions for it to load using this tool. Any help as to what this number is or how to find it is greatly appreciated. Sumsum2010·T·C 00:32, 8 November 2010 (UTC)

That's hard to say, because the sandbox was moved several times and the history is kept by page_id, not by title. But some info I've manager to to dig out:
Current title Sandbox during Revisions
Wikipedia:Sandbox March 2008 – present 330216
Wikipedia:Sandbox/Archive2 January 2008 – March 2008 33899
Wikipedia:Sandbox/Archive July 2005 – January 2008 326338
Wikipedia:Sandbox/History November 2004 – July 2005 32052
Wikipedia:Historical archive/Sandbox June 2004 – November 2004 20456
Svick (talk) 01:03, 8 November 2010 (UTC)
In general, to find the number of revisions in a page, go to the page history and find the difference between the revision with the lowest revision ID (which is usually the earliest revision or the first edit by Conversion script, but see below), and the current one. There will be a line showing the number of intermediate revisions; add two to that number and you can find out the number of revisions in a page. For example, at the time of writing, this page had 70,393 revisions. This trick will work in 99.999% of cases, but it does not work at Wikipedia:Historical archive/Sandbox because the revision ID's are out of order. Before Wikipedia was upgraded to MediaWiki 1.5 in late June 2005, if a revision was deleted and then undeleted, it would get a new ID number as if it was a brand new revision. Most of the revisions at Wikipedia:Historical archive/Sandbox were deleted in 2004 and undeleted in September and October 2007 (see the deletion logs of "Sandbox2" and "full cat litter box" and note that the linked deletion logs only go back to December 2004). IIRC, one of the old sandbox pages contained some history from June to August 2004 and October to November 2004, and the other one contained history from August 2004 to October 2004, so the history of the earliest sandbox archive page is rather disjointed! Also see Wikipedia:Database reports/Pages with the most revisions. Graham87 02:30, 8 November 2010 (UTC)
Thank you very much! After looking at the web address for the diffs. this solution makes a lot of sense to me.Sumsum2010·T·C 04:08, 8 November 2010 (UTC)

Search not working?

All possible search queries at Special:Search return zero results. Is it something I mistakenly set, or is it actually malfunctioning? As far as I know, I haven't changed any settings. —Soap— 01:55, 8 November 2010 (UTC)

It must have been a glitch; it started working after about 5 minutes. I'll leave this up in case it's symptomatic of a larger problem. —Soap— 01:59, 8 November 2010 (UTC)
02:48 <+nagios-wm> PROBLEM - LVS Lucene on search-pool1.svc.pmtpa.wmnet is CRITICAL: Connection refused  
02:49 <+nagios-wm> RECOVERY - LVS Lucene on search-pool1.svc.pmtpa.wmnet is OK: TCP OK - 0.000 second response time on port 8123  
02:52 <+nagios-wm> PROBLEM - LVS Lucene on search-pool1.svc.pmtpa.wmnet is CRITICAL: Connection refused  
02:53 <+nagios-wm> RECOVERY - LVS Lucene on search-pool1.svc.pmtpa.wmnet is OK: TCP OK - 0.000 second response time on port 8123 
02:53 -!- {Soap} [~Soap@wikipedia/soap] has joined #wikimedia-tech
02:53 < {Soap}> did the search function malfunction, or did I set something wrong?

5 Minutes seems on the money :D —TheDJ (talk • contribs) 02:10, 8 November 2010 (UTC)

Password recovery

Hi. I wanted to log into my Wikipedia account today and, realizing I forgot the password, I asked for a new one to be sent by e-mail. I can't remember which e-mail address I had associated to the Wikipedia account, so I checked all of them (I use several), but the password recovery e-mail is nowhere to be found. I looked in the spam folders too.

What can I do now? I'm worried that maybe I mistyped my e-mail (unlikely, but evidence seems to point to that). Is there any way to regain access to my account? I'm sorry if I posted this in the wrong place, btw. --85.186.76.207 (talk) 10:09, 7 November 2010 (UTC)

What is the name of your account? Otherwise I don't believe that much can be done about it. HeyMid (contributions) 10:41, 7 November 2010 (UTC)
Why is that relevant? :) --85.186.76.207 (talk) 11:16, 7 November 2010 (UTC)
Because he's trying to help you get your account back. Rehman 11:22, 7 November 2010 (UTC)
I was hoping there would be some standard procedure to follow. But if it helps, my account is User:Urzică. --85.186.76.207 (talk) 11:34, 7 November 2010 (UTC)
Special:EmailUser/Urzică says there is no e-mail associated with that account. You may be able to bug the sysadmins to reset your password. -- Cobi(t|c|b) 11:40, 7 November 2010 (UTC)
The user may have created h*s account with h*s e-mail address, but may have forgotten to confirm it. HeyMid (contributions) 11:53, 7 November 2010 (UTC)

No offence IP85.186.76.207, but, what guarantee do we have that you actually are the rightful owner of that account? For security reasons, I have left a note on the talkpage, and hopefully s/he won't reply (since its you, IP). FYI, the account made the last edit on October 30. Rehman 12:07, 7 November 2010 (UTC)

Hey wait a minute. It seems that you have created a global account way back in January 2009, with the first account created in November 2006! You are also a sysop at ro.wiki. No offence again but, what are the chances of you forgetting a four-year old password? Rehman 12:11, 7 November 2010 (UTC)

And especially considering that the username has edited as recently as yesterday (not at this wiki) suggests that this may be an impersonator. And the account is unified (SUL'ed). But Urzică may have been automatically logged in previously, and doesn't know h*s password. HeyMid (contributions) 12:21, 7 November 2010 (UTC)
You have no guarantee that I am who I claim to be, which is what makes this difficult. My password was stored on my home computer, I logged in automatically for so long that I forgot it. I am now at a different location. I'm pissed off about this whole thing, I'm aware that I might have to make a new account and lose the sysop rights and/or credibility on my home wiki (that is why I didn't ask for help there in the first place, btw). All the same, it won't be the end of the world if that happens. I'll still be able to contribute, which is what matters.
Still, I'm definitely interested to solve this problem if at all possible. We can wait for "my" answer for as long as you wish, but then what? --85.186.76.207 (talk) 12:31, 7 November 2010 (UTC)
OK, I believe you – I'm currently assuming you're the one you're pretending to be. But how did you lose your automatic login? Did you clear the cache/cookies in your web browser? Or did you simply log out by accident? Or did you switch computer? Also, as Cobi suggests, try contacting a system administrator or "sysadmin". HeyMid (contributions) 12:35, 7 November 2010 (UTC)
"but then what?", well, "I am now at a different location", so why not wait till you get back to the "earlier location" and hit the "forgot password" button? Rehman 12:38, 7 November 2010 (UTC)
And when you are logged in you can change your password by clicking on "Change password" in your preferences. Also, why didn't you think about this before you left your earlier location? And is it so severe that you actually have to edit Wikipedia right now? HeyMid (contributions) 12:44, 7 November 2010 (UTC)
Will my saved password still work if I already hit the "forgot password" button? I assumed that changes it and sends the new password by e-mail. If the old password still works, there's no problem at all. --85.186.76.207 (talk) 13:00, 7 November 2010 (UTC)
As you seemingly either don't have an e-mail address connected to your account or have forgotten to confirm your e-mail address, nothing happens if you click on the "forgot password" button. That button only works if the account has an e-mail address. HeyMid (contributions) 13:07, 7 November 2010 (UTC)
Oh, ok - looks like I panicked over nothing. Sorry for all this and thank you for your help. --85.186.76.207 (talk) 13:11, 7 November 2010 (UTC)

Just for the record - that was me indeed. Sorry again for wasting your time. --Urzică (talk) 19:07, 7 November 2010 (UTC)

Sorry for being harsh, it was for the safety of your account ;) Rehman 03:07, 9 November 2010 (UTC)

Coordinates not working?

I'm surprised no one has said anything recently, but the coordinates map seems to be broken. I click on the little globe icon and the map appears. Sometimes the red dot appears, sometimes not. The typical amalgamation of links and nearby attractions, however, never appear. Instead, "Loading" appears for a while and then changes to an almost illegible "Bad Gate" error with a link to Error. Am I missing something? --136.160.151.190 (talk) 15:34, 8 November 2010 (UTC)

IPA font update

SIL has released an expanded version of their Gentium font, which our class=IPA supports. However, they've released it under the name 'Gentium Plus'. The class should be updated to choose that over plain Gentium. (I have no idea how to do this myself.) — kwami (talk) 20:08, 8 November 2010 (UTC)

Done —TheDJ (talk • contribs) 21:45, 8 November 2010 (UTC)

Navboxes...

...are, all of a sudden, showing automatically expanded, and with no option to hide? - The Bushranger Return fireFlank speed 20:24, 8 November 2010 (UTC)

And now they're working. A few minutes ago HotCat - and cat piping when viewing Edit This Page - weren't working for a moment. Spoooky. - The Bushranger Return fireFlank speed 20:24, 8 November 2010 (UTC)
It sounds like your Javascript didn't load or encountered an error. And then it got fixed, somehow. Anomie⚔ 20:58, 8 November 2010 (UTC)
Ah, thanks. Wouldn't surprise me. I have a touch with computers...and sure isn't King Midas's. - The Bushranger Return fireFlank speed 22:01, 8 November 2010 (UTC)

http://geoiplookup.wikimedia.org/

For maybe a month, maybe a couple or so, I've noticed every time I preview a page and do some edits, and click preview again the page takes just a fraction of a bit longer to load, and it always connects to some website http://geoiplookup.wikimedia.org/ (I think, it's hard to see because it's only visible for a second). What is that? Why does it do that? Anyone know what this is about?--Brianann MacAmhlaidh (talk) 12:15, 6 November 2010 (UTC)

Actually, it isn't just previewing/editing, it's viewing articles too - like just refreshing my watchlist I'm noticing it. It's like 'Wikimedia Central' is keeping tabs on me, noting when I refresh pages and noting my location (?!). Kinda creepy.--Brianann MacAmhlaidh (talk) 12:25, 6 November 2010 (UTC)
Now that is really creepy. I don't know about you, but when I licked it, It shows my location, and I mean in coordinates! Creepy! Rehman 13:10, 6 November 2010 (UTC)
Alright now this freaked me out: I checked the coords and, its just 5 km (3.1 mi) off my dot! Rehman 13:13, 6 November 2010 (UTC)
The watchlist has had a similar service for ages. And you are being watched just as often by simply every single click you make on wikipedia (that would be called server access logs). All such "watching" is anonymized, per the privacy policy of the foundation. The geo service is included on every page in order to work around issues with caching for anonymous users. In looking into this however, it has become obvious that the geoip server does not do appropriate request caching, so for some people it will indeed download the geolocation page on EVERY single request (which is unnecessary). The ops team will now fix this I have been told. —TheDJ (talk • contribs) 13:59, 6 November 2010 (UTC)
Funny, I was just about to start a new section about this same issue. It's extremely irritating, and causing slow, slow page loads for me, typically causing edit conflicts any time I try to post on a long page. Each time there is a pause of up to ten seconds on "geoiplookup.wikimedia.org"; meanwhile my browser hangs up until it's finished loading. I notice that page source has <script type="text/javascript" src="http://geoiplookup.wikimedia.org/"></script> in it. On a little digging, I find this series of posts on the tech mailing list. Does anyone know if this is a new thing? Is there a way to make it stop? If it is necessary for Wikimedia to collect our IP addresses and log our visits, can that information be processed in the background without slowing down our page loads? Antandrus (talk) 14:18, 6 November 2010 (UTC)
Well first of all, don't use IE, that should help a lot. Also when the new Resource loader is done in a month or two, then all script loading should become a lot faster. —TheDJ (talk • contribs) 14:56, 6 November 2010 (UTC)
The easy way to stop it is to put "127.0.0.1 geoiplookup.wikimedia.org" in your hosts file. I've done it at some point because of the long delays it was causing on loading any Wikipedia pages; it's reassuring to know that I'm not missing anything as the whole thing is bloody useless (or that's my interpretation of the above discussion).—Emil J. 12:29, 9 November 2010 (UTC)
I am getting this popup beginning last week; I will add 127.0.0.1 geoiplookup.wikimedia.org to my host file (C:\WINDOWS\system32\drivers\etc) but it should not be happening. Daniel1212 (talk) 15:55, 9 February 2011 (UTC)

Table markup conversion

  Resolved
 – Smallman12q (talk) 00:43, 10 November 2010 (UTC)

I am trying to make a list of all named Maine islands. I received from the state of Maine a lengthy table in an .xls file which I think is Microsoft Excell format (I use a Mac and Microsoft is a foreign country that I don't want to visit). I need to convert the data in Microsoft .xls format to Wiki-format. How? I have been doing it manually, one table entry at a time, deleting the Microsoft formatting by hand and substituting Wiki table format--but there is just too much data for that. How do I process the whole thing in one batch?


Example of .xls format I want to convert (in edit you can see the Microsoft markup symbols):

65-250 CABBAGE ISLAND LEDGE BOOTHBAY LINCOLN U 0.5 63-032 CALDERWOOD NORTH HAVEN KNOX R 63-791 CALDWELL SAINT GEORGE KNOX R 63-794 CALDWELL (EAST MOST LT SAINT GEORGE KNOX R 63-793 CALDWELL (LITTLE) SAINT GEORGE KNOX R 63-542 CALF SAINT GEORGE KNOX R 59-177 CALF SORRENTO HANCOCK R

Example of manually reformatted Wiki-table (again--go to edit to see the substituted Wiki markup)
Registry #[1] Name Town County Owner Size Description
59-259 ADAM'S MOUNT DESERT HANCOCK private ? ?
73-029 ALDEN TOPSHAM SAGADAHOC private ? ?
63-840 ALLEN SAINT GEORGE KNOX private 450 acres seasonal
79-540 ALLEY LITTLE MARSH WASHINGTON R ? ?
59-120 ALLEY'S TRENTON HANCOCK R ? ?
59-219 ALLEY'S POINT MOUNT DESERT HANCOCK R ? ?


Thanks for your ideas ElijahBosley (talk ☞) 16:40, 8 November 2010 (UTC)

It should be very easy to do provided you have excel. Simply open in excel, copy-paste to http://excel2wiki.net and then copy-paste back to the wiki. If you can't open it, I (or someone) can do it for you. Simply upload the document somewhere (such as scribd) and post a link here.Smallman12q (talk) —Preceding undated comment added 21:01, 8 November 2010 (UTC).
WOW! It worked! Saving me and fellow editors MONTHS of tedious line-by-line transcription. There is a Barnstar for wonderfully helpful advice in your future.ElijahBosley (talk ☞) 13:50, 9 November 2010 (UTC)
Happy to help, if you need any more tech-related help post here, or to my talk page.Smallman12q (talk) 00:43, 10 November 2010 (UTC)

How to synchronization between two cas servers?

Now I have two different CAS(Central Authentication Service) servers. Browser to access web servers may be dispatched to these two CAS servers. How can I do to achieve, so no matter where redirection CAS servers can be authenticated. —Preceding unsigned comment added by Vincen koo (talk • contribs) 02:30, 10 November 2010 (UTC)

This page is for dealing with technical issues related to Wikipedia. Try the computing reference desk. Graham87 06:05, 10 November 2010 (UTC)

To Google Reader users: You may be missing items from your watchlist feed.

I want to apologize for my English in advance, as it's not my native language. I also want to apologize in case this isn't the right place to post this or if this issue has being already reported.

I'm following my watchlist updates through RSS feeds both in English and Spanish Wikipedia, but I've just realized that Google Reader was missing some of the updates. As "My watchlist" Wikipedia RSS feeds don't validate, I thought that, perhaps, changing to Atom I'd solve the problem. After all, "My watchlist" Wikipedia Atom feeds do validate. But then I found that Atom versions also miss items.

I have started a thread on Google Reader Help Forum to report this issue. I 'm "nornand" there.

Therefore, I'm tracking currently four watchlist feeds (RSS-Spanish, Atom-Spanish, RSS-English and Atom-English) and all of them have missed items in Google Reader. In my humble opinion, this is a serious problem because many people rely on GR to follow their feeds and Wikipedia editors aren't an exception for sure. When you add an article to your watchlist you do it for good reason (to control the quality of the updates, protect it against vandalism, etc.) so if your feed reader fails you'd better take precautions. In other words, I don't know if this issue is going to be solved some day or how big is its scope (my feeds are affected in two different languages), but I feel that it had to be reported.

It wold be nice if someone using Google Reader to follow his watchlist checked if he has the problem too.

Thank you and, again, sorry for my English and for the long post. I wait for your feedback.--Canyq (talk) 03:18, 8 November 2010 (UTC)

I think this is the same bug as T25686. Svick (talk) 00:47, 10 November 2010 (UTC)
Hummm, yes, it may be the same bug, beacause although Atom feed validates, a warning about "entries with the same id" is given. As it is said in the bug page, Wikipedia feeds should use an unique ID instead of elements like article titles, which may appear in several entries. At this moment, my watchlist RSS feeds are a downright mess in terms of losing items, so my recommendation is using Atom versions although, as I have already said, they also lose some items. Anyway, I insist, this is a serious problem as it involves one of the most widespread feed readers, therefore either Google Reader or Wikipedia should try to fix it. But until that moment, my advice is that if you use Google Reader and are truly concerned about your watchlist, check updates manually in the webpage or try another feed reader. And if you want to keep on using Google Reader, at least, use Atom instead of RSS feeds (although you'll keep on losing entries). --Canyq (talk) 17:20, 10 November 2010 (UTC)

Can I opt out of geonotices?

I'm currently seeing a JS error when I load my watchlist, coming from the geonotice.js module, which I assume is linked to WP:Geonotice. As a result the shortcut links to section headers aren't working. I've seen this happen before and I see a thread above about "geoiplookup". How do I completely opt out of any geonotice processing? If there's something of compelling interest to my "area" I will either already know about it or not care. I don't want the extra processing and I especially don't want to lose functionality because of some obscure problem. I don't want geonotice.js to load or execute on my system, period. Is there a way to opt out? (Yes I refreshed my cache, yes I use IE, no I don't care - I don't want the (non-)functionality) Thanks! Franamax (talk) 21:19, 8 November 2010 (UTC)

You can't opt out as far as I know. Not right now anyway. —TheDJ (talk • contribs) 21:30, 8 November 2010 (UTC)
Thanks (I guess... :) Is there a place to discuss it? If it's only for watchlists it should be easily gadget-able. I'll try WT:Geonotice I guess. Franamax (talk) 23:29, 8 November 2010 (UTC)
The WP:Geonotice. say "It can permanently be hidden by adding "#WN_GEON { display: none; }" to your CSS file." not tried this to see if it works or not.--Salix (talk): 12:35, 9 November 2010 (UTC)
But hidden != disabling. Hidden will still run the code and load the resource in question. —TheDJ (talk • contribs) 15:33, 9 November 2010 (UTC)
Precisely. Pages take long enough to load as it is. I'm sayin' don't bother to figure out where I am, I already know and I don't need or want the help. Hiding the results doesn't stop the processing and it doesn't fix the JS error I'm still seeing every time I check my watchlist. Franamax (talk) 23:42, 10 November 2010 (UTC)
Tools such as adblock plus for FireFox work great, Ive opt'ed out of several really annoying javascript addons. ΔT The only constant 15:47, 9 November 2010 (UTC)

Strange tripping of the spam edit filter

Here is a good one. Take the following div statement: <div style="height: 90px; overflow: auto">, now switch the order of the "height" and "overflow" statements, and try to save it. In other words, try to write <div style="overflow: auto; hight: 90px">, but with 'height' correctly spelled. Currently, this trips the spam filter, and it won't let you save it! I have no idea what is going on with this. I was able to circumvent this problem by swapping the order, but it seems very strange that a div style statement would have anything to do with the external link filter. Thanks! Plastikspork ―Œ(talk) 22:45, 9 November 2010 (UTC)

The filter says the problem is with the phrase "overflow: auto; hight:", again with things correctly spelled. I can't seem to find any such entry on either the local or global blacklists. Hersfold (t/a/c) 23:30, 9 November 2010 (UTC)
And to clarify, this is a spam blacklist problem, and nothing to do with the Special:EditFilter. Hersfold (t/a/c) 23:31, 9 November 2010 (UTC)
I don't think it's the "standard" mw:Extension:SpamBlacklist blacklist either, even though it's giving the same message. It triggers even if you enter <nowiki>overflow: auto; height:</nowiki>. Anomie⚔ 00:10, 10 November 2010 (UTC)
Google helped. MediaWiki talk:Spam-whitelist/Archives/2008/02#overflow: auto.3B height:. I'd have to agree with Isaac Dupree there. Anomie⚔ 00:32, 10 November 2010 (UTC)
Weird... I was apparently the one who brought all this up in the first place. Funny how that happens. Hersfold (t/a/c) 07:38, 10 November 2010 (UTC)

To expand on "Blocking of this string occurs in the server configurations": InitialiseSettings.php.txt see the line '/overflow\s*:\s*auto\s*;\s*height\s*:|<div[^>]*font-size[^>]*font-color:\s*transparent[^>]*>/i'] under wgSpamRegex. Also see $wgSpamRegex, this is the raw method of inserting spam trips, without extensions. --Splarka (rant) 09:39, 10 November 2010 (UTC)

Search engine feature questions

I'm currently surveying Wikipedia's coverage of ice hockey, and have been messing around with searches and search links.

What I need to know is:

1. Is there a way to specify in the search itself how many results to display per page? (I'd like to set it at 500 or higher).


2. Is there a way to specify punctuation marks such as parentheses?

For example, this search:

doesn't work.

I'd like to not see results with "(ice hockey)" at the end of the title like this one: Andy Murray (ice hockey)

The Transhumanist 23:04, 9 November 2010 (UTC)

  1. The hard limit is 500 results, mainly because setting it higher can produce (sometimes unintentional) strain on our search backend during peak times.
  2. No, punctuation marks and wikitext are by default ignored. I think it would be useful to have a raw-text search where both would be searchable, but as far as I know, no-one is working on it. --rainman (talk) 00:08, 10 November 2010 (UTC)
At Special:Preferences under "Search options" you can choose Hits per page. On a search results page you can click a number at the bottom to change the number of hits for that search, and to see what the url looks like. Afterwards you can manually change the number in the url. I don't know whether the number can be controlled with a wikilink to Special:Search. Help:Searching#Specialist searches links Grep which can search punctuation in page titles (see regexp), but cannot search page content. I don't know whether it can be used to exclude "(ice hockey)" like in your example, or whether it can give many search hits. PrimeHunter (talk) 03:15, 10 November 2010 (UTC)
(Please feel free to answer between my responses below...)
The "Hits per page" preferences tip helped me a lot. Thank you. I also changed "Lines per hit" to 1, but it didn't have any effect. What is it supposed to do? I'd like to get results that list only the page names.


Is there any way to get the results in wikicode? I'd like the results to include link brackets.


Grep is great. It returns page names without the extra contextual stuff. And it gives many hits (though I haven't explored its limits yet). I never knew it existed. Thank you!


When I copy/paste results from Grep, there's an empty line between each one. Is there a way to get Grep results without these?


Is there an easy way to remove empty lines from a page?


I can't figure out how to use Grep to exclude strings. Eg., return titles with "Albania" in them but not those with "Superliga" in them.


But you can exclude characters from a string like the "(" from "(ice hockey)". For this I used "[^\(]ice hockey". It returns titles with "ice hockey" and not those with "(ice hockey"


You have been very helpful. Again, thank you!The Transhumanist 01:00, 11 November 2010 (UTC)
It appears from my experiments and Help:Wikipedia: The Missing Manual/Customizing Wikipedia/Customizing with Preferences#Search (which I found by searching "Lines per hit") that "Lines per hit" has no effect. I don't have answers to the other questions. PrimeHunter (talk) 01:05, 11 November 2010 (UTC)

Adding CSS code to one's personal skin page based on what server they're on

I've recently switched over to using the secure server to edit from, but I keep thinking that somehow I've managed to log onto the unsecure server by mistake and I'm on that one instead. To help reduce the panic, I'm trying to add some CSS code to my Monobook page to turn the backgrounds of pages a red tint (rather than the normal blue) if I'm on the secure server. Naturally, though, there isn't a whole lot of conditional logic in CSS (what there is seems limited to making Internet Explorer work) and my attempt to use parser functions also failed, likely because .css pages don't parse wikimarkup. Does anyone have any ideas for making this work, or is this an exercise in futility? Hersfold (t/a/c) 23:25, 9 November 2010 (UTC)

I think probably, yes. You could achieve a similar result with custom js though I think. The same result if, for example, you modify the <body> tag directly. OrangeDog (τ • ε) 23:38, 9 November 2010 (UTC)
You would need to do it from the JS said and have it add/modify the css that you want. Peachey88 (T · C) 23:40, 9 November 2010 (UTC)
Not being terribly familiar with javascript, how exactly would I do that? Hersfold (t/a/c) 00:18, 10 November 2010 (UTC)
This should do the trick —TheDJ (talk • contribs) 01:31, 10 November 2010 (UTC)
if( wgServer.indexOf( "https://secure.wikimedia.org" ) == 0 ) {
importStylesheet( "User:Hersfold/secure.css" );
}
Login on the normal and secure servers is separate, so if you log out on the normal server, you'll be able to see the difference because some buttons go missing. Ucucha 01:04, 10 November 2010 (UTC)
(ec) Or, as OrangeDog said, you could use JS: this makes the text of User:Ucucha/sandbox appear in italics on the secure server but in normal font on the non-secure server. Ucucha 01:33, 10 November 2010 (UTC)
The javascript worked - thanks much, everyone! Hersfold (t/a/c) 07:36, 10 November 2010 (UTC)

Strange glitch has been bothering me lately.

The standard toolbar (The one that goes Bold, Italic to <ref /ref>, {{ Cite }}) occasionally loses all the buttons to the right of Horizontal line, and I need to refresh the edit page to get them back.--occono (talk) 03:26, 2 November 2010 (UTC)

Perhaps it's a browser-related issue? Do you have a recent version?—RJH (talk) 19:20, 3 November 2010 (UTC)
It has affected me across mulitple computers and browsers. I thought getting rid of some gadgets and moving to the new skin might have fixed it, but it's been happening again.
It's still happening. --occono (talk) 03:00, 5 November 2010 (UTC)
I've copied this from the Archive because it's still happening.--occono (talk) 01:42, 11 November 2010 (UTC)

Archiving issue

Hey, can somebody figure out why Talk:Risks to civilization, humans and planet Earth won't archive properly? I tried messing with Miszabot but suspect it was in vain. —K. the Surveyor (talk) 04:09, 11 November 2010 (UTC)

yeah, it's set to archive every 365 days. I'll set it to two weeks instead - give it a day or so to kick in. --Ludwigs2 06:28, 11 November 2010 (UTC)
i did that...just a short while ago. it had been set i think to 90 days or something. and we have posts from 2006. well anyway, hopefully it will work now... —K. the Surveyor (talk) 06:35, 11 November 2010 (UTC)

New script: Print dialog

Tonight I wrote a script that I think many will welcome. It is called Print dialog and it changes the "Printable version" link into a "Print page" link. When clicked, you get a dialog that allows you to make some choices. You can print everything which is normally hidden (for instance interface, metadata etc), but also force background to disappear, make all text black, hide references and hide images. It's in very early testing, and only tested with Safari, and possibly it might never work very well on IE, but IE sucks, so who cares :D. Have fun testing it, and let me know what you think. Simply add importScript('User:TheDJ/printdialog.js'); to your script page. —TheDJ (talk • contribs) 03:36, 11 November 2010 (UTC)

Screenshot of the dialog earlier in the night, and a print to pdf version of printing with all interface elements enabled: HERE. —TheDJ (talk • contribs) 03:40, 11 November 2010 (UTC)
I have now updated the documentation page with explanations and examples. —TheDJ (talk • contribs) 20:38, 11 November 2010 (UTC)

Large reference list breaking mediawiki software

I hope I'm putting this in the right place, but having tried everything I can think of and asking questions elsewhere, I think this is a problem with mediawiki software. On List of civil parishes in Somerset there is an extremely large list of references which seems to be too many for the software to display - can anyone suggest any solutions?— Rod talk 09:21, 11 November 2010 (UTC)

Got the references to show up, although a bunch of them still are not displaying correctly (just displaying "cite web". I think the problem is the page is transcluding too many templates, which would explain why the cite templates aren't expanding properly, and also why the references didn't show up before (I switched the template {{reflist}} for a direct use of <references />), and why the Featured list template isn't working. Probably a setting somewhere to limit the number of templates any one page can transclude. - Kingpin13 (talk) 09:42, 11 November 2010 (UTC)
Some possible further reading: Wikipedia:Template limits, maybe Special:ExpandTemplates would fix the problem - I don't know - Kingpin13 (talk) 09:46, 11 November 2010 (UTC)
Thanks for trying. I've looked at the pages you pointed me to but they are beyond my technical capabilities & if I start fiddling I will probably break it.— Rod talk 10:00, 11 November 2010 (UTC)
I'm trying to remove some of the deadlink & similar templates. It all worked fine when it was promoted to FL but ironically I think it was the additional of the star template which broke it.— Rod talk 10:23, 11 November 2010 (UTC)
Interesting, there's a Category:Pages where template include size is exceeded which notes other articles with a similar problem.  pablo 10:41, 11 November 2010 (UTC)
The page was exceeding the "Post-expand include size" (as explained on Wikipedia:Template limits), which is to do with the number/size of templates rather than references themselves. As an immediate fix I have switched to the {{vcite web}} etc. templates, which changes the citation style but does mean they all show as there is less code. I have also replaced {{convert}} and {{centre}}. mattbr 11:36, 11 November 2010 (UTC)
Thank you. I will not claim to understand all of that but I can see that it has worked. Would it be worth some sort of note on the talk page of the 1000+ articles in the category given above to explain the problem?— Rod talk 11:55, 11 November 2010 (UTC)
You're welcome. I don't fully understand it either but it has to do with the size and complexity of the templates used. The reason for the page breaking may have been due to changes to the templates used rather than any action on the page itself. You might want to consider manually formatting the references or using a shortened system as described at Wikipedia:Citing sources#Shortened footnotes. Don't think there is an alternative to using {{coord}}? Many of the pages in that category are outside article space, but it might be worthwhile on the articles. mattbr 16:40, 11 November 2010 (UTC)

Chembox too long

Comments? Wikipedia talk:Chemical infobox#Chembox too long
--Wickey-nl (talk) 14:33, 11 November 2010 (UTC)

Any automated tools exist to check on the creation of unsourced NEW articles by a user?

Do any automated tools exist to check on the creation of unsourced NEW articles by a user?

I add a lot of citations to Wikipedia, as well as tag {{citation needed}} to a lot of unsourced claims. From time to time I find a user who regularly creates NEW articles with no sources whatsover. Sometimes, the editor cites some of her/his NEW articles, while making exclusively unsourced claims in others. Or the editor sources a higher percentage of articles nowadays versus, say, a couple of years ago. It would be helpful to have an automated way of examining the NEW article history, sources or not, of a user--maybe get the percentage with sources and without over time. Anyone ever create a tool to do this? Cheers. N2e (talk) 19:25, 11 November 2010 (UTC)

IF you can define what a "sourced" article is vs unsourced, it shouldnt be that hard. ΔT The only constant 19:43, 11 November 2010 (UTC)
The basic criteria that would be most useful to look for would be a "==References==" section with at least one set of "<ref> ... </ref> reftags in the article. An article with "==External links==" might be considered to have some reliable source support, but it could not really be said to have the inline citations that are preferred by WP policy: "The policy on sourcing is Verifiability, which requires inline citations for any material challenged or likely to be challenged, and for all quotations." (quoted from WP:CS). In my view, at least one inline citation is required on a NEW article because any NEW article with no sources is likely to be challenged.
Of course variations that remain compliant with Wikisyntax could also occur, such as: "== References ==" (with one or more spaces). What do others think? N2e (talk) 21:06, 11 November 2010 (UTC)
I've seen a lot of cases where a new user will put a direct external link inline with their text: [http://...] That will not generate anything in a References or External Links section, but is still referenced, just not formatted properly. Perhaps that is something that would still be useful to know though, as it is an opportunity to educate a new contributor before they get template-bombed off the project. So perhaps for the bot, an article that does not contain the string "<ref>"? Of course there's also the fact that a lot of new articles don't have any refs on their first edit, but do shortly thereafter (I'm guilty of that), so flagging for the N flag and no refs, might have too many false positives ArakunemTalk 21:29, 11 November 2010 (UTC)


Few new articles these days should go for more than a day or two without an external link. Just pop a deletion tag onto the article and the author should wake up and clean it up a bit. To automate the process, just write something to scan the new articles stream, hold for a day or two, and list those that haven't ever contained any external links of any kind.

You can easily spot an external link; it's a string starting with a scheme (http, ftp, or whatever) followed by ":" and some stuff. Simply this to "http://" which accounts for nearly every external link. At this point a human is asked to look at the link-free articles. If they contain a clearly verifiable source (ie: the human looks at the article and is able to obtain and read the source and verify it) then it's removed from the list. If not, then it's on the suspect new articles list. --TS 22:09, 11 November 2010 (UTC)

Well, apparently not. In only a little looking on a single user, I found two new articles from August 2010 (three months ago) that are unsourced: Liogorytes and Philanthus ventilabris.
But that wasn't my question. I'm just wondering if there is an automated tool that would allow a search of a specific user to see what other articles that User has created, over the past few years, that were unsourced then and still are today. Cheers. N2e (talk) 22:56, 11 November 2010 (UTC)

user talk edit notice

User talk:Jimbo Wales/Editnotice is appearing when editing User talk:Jimbo Wales. Is that supposed to happen? --John Vandenberg (chat) 03:10, 12 November 2010 (UTC)

Yes, technically that should happen. I've deleted the talk page of the editnotice. Editnotices should not have such talk pages, it breaks stuff. —TheDJ (talk • contribs) 03:21, 12 November 2010 (UTC)

Need template help

What am I doing wrong here? Thanks, Access Denied – talk to me 04:31, 12 November 2010 (UTC)

I think you need {{#if:{{{4|}}}|....}} rather than {{#if:{{{4}}}|....}} having a pipe in the parameter gives the empty string if the parameter is not defined, without the pipe the if clause evaluates to true when not parameter is set. See mw:Help:Parser functions in templates.--Salix (talk): 06:42, 12 November 2010 (UTC)

User contributions to a specific page

Is there a URL available for viewing a user's contributions to a specific page? Thanks, 28bytes (talk) 18:07, 11 November 2010 (UTC)

Maybe something at Help:Page history #Searching and exporting histories. DMacks (talk) 22:51, 11 November 2010 (UTC)
You can use this tool. Graham87 02:21, 12 November 2010 (UTC)
Also see Wikipedia:Village pump (technical)/Archive 62#Finding all edits a user made to a certain page. Graham87 02:28, 12 November 2010 (UTC)
Thanks both for the links. I had hoped that there would be a parameter I could pass to the Special:Contributions URL; since it lets you limit the results by namespace, I thought it might let you limit them by page name also. As a related question, is there a list of valid parameters to the Special:Contributions URL anyplace? Thanks, 28bytes (talk) 18:35, 12 November 2010 (UTC)

Main Donate link doesnt work

Click on the Donate link at the left and it goes to [1]. Didnt know where or who to report it to. --Matt57 (talk•contribs) 13:16, 12 November 2010 (UTC)

Your link failed for me earlier but works now. PrimeHunter (talk) 13:51, 12 November 2010 (UTC)
WMF employees are actively preparing for the donation drive, and for the past days, new software for that has been deployed multiple times a day. You probably hit exactly one of those deployments. —TheDJ (talk • contribs) 14:10, 12 November 2010 (UTC)
Thanks for the report - I think it's fixed! :) Philippe Beaudette, Wikimedia Foundation (talk) 17:35, 12 November 2010 (UTC)

Retrieval of passwords/committed identity

Would one of the experts please comment at this thread on the Help Desk. In brief, if you have a "committed identity" is it possible for a lost password to be retrieved if you no longer have access to the e-mail account in your user preferences? Thanks. – ukexpat (talk) 15:23, 12 November 2010 (UTC)

No and Yes. You do not get to retrieve the lost password, but a sysadmin might be able to set an updated email address, if you can be positively identified with your committed identity, if I understand correctly. This is usually only done for contributors with a large edit history, because it is a rather laborious process. —TheDJ (talk • contribs) 16:00, 12 November 2010 (UTC)
Thanks for the clarification. – ukexpat (talk) 16:13, 12 November 2010 (UTC)
To clarify a minor point: the only reason for the "committed identity" in the first place is to help in convincing the sysadmins that the random person asking for access to a longstanding account is really the owner of that account and not just someone engaged in social engineering. If you can convince them of that in some other way, there is no need for a "committed identity". Anomie⚔ 18:12, 12 November 2010 (UTC)
Wikipedia:Help desk and Wikipedia:New contributors' help page get many posts from users who say they forgot their password or "the password no longer works", and they haven't stored an email address or no longer have access to the email account or don't receive a new password when it's requested at login. If it doesn't help to remind them of upper/lower case distinction, typing carefully, and checking their spam folder or ISP for password mails, then they are normally told to create a new account and possibly make links between the two accounts. It's usually users with relatively few edits and no extra permissions. Is there advice somewhere on when and how to tell them to consider bugging the sysadmins to try and gain access to the existing account? I don't know anything about when the sysadmins will bother looking into it or what they require as proof of identity, but I guess they are busy and prefer to avoid dealing with this. PrimeHunter (talk) 23:04, 12 November 2010 (UTC)

Fundraiser: no donate form for readers in Switzerland

Hi. As you may see at this donation page for readers in Switzerland or its capture at ImageShack, there is no obvious nor visible form or link to donate. The only way to donate provided by the Swiss chapter Wikimedia CH is the old-fashioned bank check. There is no online form to donate, no Paypal, and so forth. The link to donate to the WMF is in small, at the bottom of the page.

This is horrible, I would never want to go trough all the troubles to make a donation with a bank check. This is Internet, in fall 2010. Not the 90's.

Just what do we expect users to do? It's great to allow local chapters to participate in the Fundraiser. But they should follow a framework, or a set of guidelines, to ensure they provide a reasonably usable and efficient donation page. I suggest three major guidelines:

  1. The chapter landing page must be translated into the official language(s) of the country, plus English. In addition, translation into every major languages spoken in the country would be appreciated.
  2. There must be an explicit and obvious way to donate online, via a credit card or Paypal.
  3. Let the user choose if he wants to donate to the local chapter, or the WMF. Both should be as obvious, and one option should not be voluntarily set aside.

Kind regards, Dodoïste (talk) 19:55, 12 November 2010 (UTC)

That page is owned by the Swiss Wiki Chapter. I would suggest passing these suggestions on to them. 216.38.130.163 (talk) 20:05, 12 November 2010 (UTC)
The Swiss have total control over their own landing page. They have told us that no one donates online, and individuals would prefer to give through check. While chapters control their pages, we have suggested best practices and how to optimize their donation pages. Please relate your concerns directly to the Swiss Chapter --Deniz (WMF) (talk) 20:08, 12 November 2010 (UTC)
Probably specifically, this should be brought up at http://meta.wikimedia.org/wiki/Talk:Fundraising_2010 and at http://www.wikimedia.ch/Main_Page--108.27.105.73 (talk) 20:13, 12 November 2010 (UTC)
OK, I'm discussing this issue with the Swiss community as of now. Thanks. :-) Dodoïste (talk) 21:17, 12 November 2010 (UTC)

Navbox collapsible collapsed

Some time in the past 48 hours or so, "navbox collapsible collapsed" (which I use for {{tip of the day}} and {{did you know}} on my talk page) stopped functioning properly and is no longer displaying show and hide buttons. I don't understand what happened. Anyone here know? – Athaenara ✉ 20:12, 12 November 2010 (UTC)

You might have a broken script cached. Try to bypass your browsercache. —TheDJ (talk • contribs) 20:25, 12 November 2010 (UTC)
I've been clearing my browser cache 3 or 4 times a day for more than a year as a matter of routine; this problem didn't occur before this mid-week. – Athaenara ✉ 21:25, 12 November 2010 (UTC)
Do you have any Gadgets enabled ? —TheDJ (talk • contribs) 21:44, 12 November 2010 (UTC)
I think not, as I don't know what "Gadgets" might be in this context and have never installed or enabled any (I use Safari). – Athaenara ✉ 22:07, 12 November 2010 (UTC)
"Gadgets" is a tab at Special:Preferences. Have you tried logging out? Is Javascript enabled and working elsewhere? (see Help:Collapsing#Limitations). Can you try another browser? PrimeHunter (talk) 00:41, 13 November 2010 (UTC)
Ah, those gadgets (they haven't crossed my mind since last year's fundraising bannerspam)! Just one gadget is enabled in my preferences (Change the "new section" tab text to instead display the much narrower "+") and has been unchanged at least since 2007. Javascript is installed and has been working as expected in most respects. I don't think I want to change browsers solely because of one out-of-whack feature. – Athaenara ✉ 02:11, 13 November 2010 (UTC)
I meant whether you can easily test another browser (I have 4 installed but that's probably unusual and Safari isn't among them). I didn't suggest a permanent browser change. I guess it's a Javascript issue in your Safari. The show and hide buttons disappear if Javascript is disabled but there are probably other Javascript problems that can cause it. PrimeHunter (talk) 02:26, 13 November 2010 (UTC)
OK, I'll give it some more thought, and thanks, both of you, for all the fresh ideas. – Athaenara ✉ 02:48, 13 November 2010 (UTC)

Strange fundraising appeal display

 
Exhibit A

In my browser (Firefox 2.0.0.20), the text of the fundraising appeal displays at a huge size and overwrites a lot of other stuff, and there's no obvious place to click on to dismiss the box... AnonMoos (talk) 22:07, 11 November 2010 (UTC)

What skin is that? Kaldari (talk) 22:17, 11 November 2010 (UTC)
"Classic". Finally figured out how to dismiss the box, after reading the thread several sections above... AnonMoos (talk) 22:47, 11 November 2010 (UTC)
Weird. It looks fine to me in Classic skin in Firefox. Do you have any custom CSS? Kaldari (talk) 23:04, 11 November 2010 (UTC)
No. AnonMoos (talk) 23:21, 11 November 2010 (UTC)
Firefox 2.0 is very old. Please upgrade to 3.6. — Edokter • Talk • 23:17, 11 November 2010 (UTC)
It's the most-recently released version that works with the operating system I use, and I don't think it's so antiquarian that Wikipedia developers should completely ignore it. AnonMoos (talk) 23:21, 11 November 2010 (UTC)
What OS are you using? And on what page are you seeing the errors? — Edokter • Talk • 23:31, 11 November 2010 (UTC)
Firefox 3.0 (released 2008) dropped support for Windows NT4/98/ME which were End-of-Life'd in 2006. It's missing some important things like inline-block (use -moz-inline-box) and usage share for 2.0 is 0.38% of all browsers. Honestly, these people should upgrade or switch to Linux before their boxes join a botnet. — Dispenser 04:58, 12 November 2010 (UTC)
I have just reenabled the "Hide fundraiser banner"-gadget, in anticipation of more annoyed editors, starting tomorrow. :D —TheDJ (talk • contribs) 23:54, 11 November 2010 (UTC)
I added a script to the gadget that will avoid loading the banner code alltogether (the request is blocked). The advantage is that if you have Google Chrome or Safari, the image won't even be downloaded for instance and the code will not not be inserted into the DOM (better and more efficient than hiding). —TheDJ (talk • contribs) 18:56, 12 November 2010 (UTC)

Cannot Wikipedia be more be careful with his human resources? Instead of irritating with repeated fundraiser banners, people who are logged-in can get the message once, with the fundraiser banner blocked by default.--Wickey-nl (talk) 11:35, 13 November 2010 (UTC)

How does the "new section" link really work?

Often I'll be responding to a question on the Help Desk or one of the Ref Desks and it will be the bottom section on the page (i.e. the last question). When I do this, I occasionally get hit by a edit conflict message. When I check on what the conflict is/was, it will be someone adding a new section/question. Should I assume that this means that when you click on the + sign, it is actually editing the entire page?

Also, I just posted a question on the Computers section of the Ref Desk (it's about Mac computers and if any tech people here have an answer, that would be great). It's on my watchlist. When I saw that someone had edited the section that my question is in (noted by the grey italics that is in the watchlist listing), I thought "Sweet! An answer for my question" and popped right on to the page. Only once I got there, I realized that my question was not answered. Instead I found that someone had added a question below mine. So did that person then edit my question to add one of their own? Or did they use the 'add section' link? Dismas|(talk) 17:29, 12 November 2010 (UTC)

  1. I may be wrong, but I think editing a section merges the section you edited into the version of the page that you started from, and then saves that whole page as if you had edited the whole thing. Then the edit conflict resolver tries to merge your changes with the current version of the article, without having any idea whether you or the other guy added a section, edited a section, or edited the whole page.
  2. Someone edited your section to add their question below yours instead of using the "new section" button. And then didn't correct the edit summary. I personally hate when people do that, but there are too many of them to try to fight it. Anomie⚔ 18:08, 12 November 2010 (UTC)
See Wikipedia:Use the new section button. Graham87 04:05, 13 November 2010 (UTC)
Great! Thanks both of you! I too hate it when people don't use the new section button! Dismas|(talk) 16:04, 13 November 2010 (UTC)

That banner is doing funny things

I already dismissed the the "personal appeal" banner a couple days ago, after reading it, however, it is somehow "there". What I mean, is that when I try to use my cursor to select the title of a page, the title does not appear highlighted, and when I copy, paste, the text on the banner (a personal appeal...) appears where I paste it. Brambleclawx 21:31, 12 November 2010 (UTC)

What browser + version + Operating System are you using ? —TheDJ (talk • contribs) 21:47, 12 November 2010 (UTC)
IE7. OS: Vista. Brambleclawx 02:35, 14 November 2010 (UTC)

Links to sections

Did you notice that links to sections, such as MOS:QUOTE, are now linking several lines above the intended section WP:Manual of Style#Quotations (that link has the same problem)? Perhaps the fundraising banner causes that. Firefox version 3.6.12, Windows XP Professional Version 2002 Service Pack 3. Art LaPella (talk) 02:24, 13 November 2010 (UTC)

Yes, it's the banner. It happens to me in all 4 tested browsers: Firefox, IE, Opera, Chrome. It works if the banner is removed with the x in the corner. It only happens when I come from another page. If I then click the browser address bar and Enter then it goes to the right place. The TOC always goes to the right place. PrimeHunter (talk) 02:40, 13 November 2010 (UTC)
Hmm, that is a tad annoying. Should be possible to check for Anchors and adopt the scroll position. I'll leave a note for the tech team that works on that. —TheDJ (talk • contribs) 12:25, 13 November 2010 (UTC)
I'm not sure that has anything to do with the fundraising banner specifically, I've noticed the same problem generally when a script or unsized image load changes the layout of the page after the browser has jumped to the anchor in question. Anomie⚔ 23:28, 13 November 2010 (UTC)
Yes, but that doesn't make this case less annoying. With a 172px high banner.. —TheDJ (talk • contribs) 23:51, 13 November 2010 (UTC)
Yes, I've found this quite annoying as well. Killiondude (talk) 00:13, 14 November 2010 (UTC)

Hotcat

Hi, I would like to execute the Hotcat script which I installed; how can I do this? Mar4d (talk) 08:23, 13 November 2010 (UTC)

It will show up at the bottom of articles. You shouldn't have to do anything else. —TheDJ (talk • contribs) 12:27, 13 November 2010 (UTC)
Sorry...I meant to say that I want to deactivate it. How can I do that? Mar4d (talk) 04:33, 14 November 2010 (UTC)
My preferences>Gadgets>HotCat Choyoołʼįįhí:Seb az86556 > haneʼ 04:41, 14 November 2010 (UTC)

Bloody banner

I've dismissed it but it won't go away, or rather it will until I return to my watchlist. It also gives me a run error every time I go to my watchlist (line 39 - paypal something) as does the appeal itself. I've cleared my cache & still get this on both IE & Firefox. Johnbod (talk) 16:26, 13 November 2010 (UTC)

Hey Johnbod. I'm looking into this now I haven't been able to replicate it yet or see what the issue is. I assume you have cookies enabled? (the dismiss banner works on a 1 week per project cookie) Jalexander--WMF 21:01, 13 November 2010 (UTC)
I did wipe all my cookies & the problem persisted. But now it is over & in fact I don't see the banner at all. So crisis over, if this was just something on my machine - Firefox had crashed after the banner first appeared but before the problem started, which may account for it. I should have said that before probably. Thanks for your efforts! Johnbod (talk) 00:37, 14 November 2010 (UTC)
We have a gadget to handle this, go into your preferences and on the gadgets tab it can be found under "Browsing gadgets". Peachey88 (T · C) 02:23, 14 November 2010 (UTC)

I can't view any videos on Wikipedia, why?

I have been reading Wikipedia for almost as long as it has been on the web. I have never been able to view any videos or listen to any songs on it. I have various players installed on my current laptop that has Vista installed and I have no problem viewing any videos or listening to any songs on any other websites. I do not understand why nothing seems to work for me on Wikipedia. I even clicked on the 'run activex' (or whatever it is) and it still would not run. Nothing happens. Animations, such as geological animations have worked fine on Wikipedia. I do have cookies limited, but that has never prevented me from viewing things on other websites. Can someone please help me??? Thanks.Mylittlezach (talk) 02:22, 14 November 2010 (UTC)

See Wikipedia:Media help. PrimeHunter (talk) 03:38, 14 November 2010 (UTC)

Stable wormhole between two talk: pages

Does anyone think User:NYKevin/interface is potentially useful anywhere? I can't help but think it must be useful somewhere... but I can't quite figure out where. --NYKevin @007, i.e. 23:09, 10 November 2010 (UTC)

Yet another frighteningly complex workaround for our feature-poor wiki-based discussion interface. We should be able to easily display conversations like talk page discussions on both talk pages without doing any fancy transclusion. Dcoetzee 10:02, 14 November 2010 (UTC)
So should I transwiki it to Wikiversity or something? --NYKevin @302, i.e. 06:14, 15 November 2010 (UTC)

On-Line Encyclopedia of Integer Sequences links

I've come across a few pages that deal with numbers and sequences. One specific example is at Powerful number. Without looking in-depth, the first link to A001694 leads to a wiki page, which says that the sequence was deleted from the wiki and added to the main OEIS database. I found that on that page if you delete the /wiki element of the URL at that point, it goes right to the desired sequence. I can't be sure if all the OEIS references end up in the same fashion, so I don't feel at liberty to change the templates as listed at Talk:On-Line_Encyclopedia_of_Integer_Sequences#OEIS_linking_syntax_help and possibly elsewhere. And I didn't think that this comment quite belongs on that talk page. D. F. Schmidt (talk) 19:48, 13 November 2010 (UTC)

Looking at the template {{OEIS}} it seems to depend on the meta interwiki map – e.g. [[OEIS:A000001]] generates OEIS:A000001 but, from what you said, should generate OEIS:A000001. That will require a Meta admin to change. Rambo's Revenge (talk) 20:05, 13 November 2010 (UTC)
A bad interwiki map change [2] took effect today and broke nearly all OEIS links in Wikipedia. I have posted a request to change it back at meta:Talk:Interwiki map#Requests for updates. PrimeHunter (talk) 21:41, 13 November 2010 (UTC)
Even if the change should be reverted quickly at meta:Interwiki map it can apparently take months before the interwiki database is updated so the change takes effect. Is there a way to speed up the interwiki fix at the English Wikipedia? If not then I think we should edit {{OEIS}}, {{OEIS2C}} and {{Number of relations}} to use http://oeis.org/ instead of relying on the OEIS prefix. These are apparently the only templates using the prefix. Is there a way to find articles using the OEIS prefix directly and not via a template? PrimeHunter (talk) 22:39, 13 November 2010 (UTC)
I've temporarily changed the templates to use http://oeis.org/ rather than the interwiki.--Salix (talk): 23:42, 13 November 2010 (UTC)
interwiki map now changed. Don't know how long it will take for the change to take effect.--Salix (talk): 00:14, 14 November 2010 (UTC)
Good. It appears from [3] it took 3 months last time. Thanks for updating the templates. PrimeHunter (talk) 00:22, 14 November 2010 (UTC)
Per my request, jeluf synced it again today. —TheDJ (talk • contribs) 02:04, 14 November 2010 (UTC)
After verifying that the interwiki map is working as intended, I've reverted the templates to use that again. Just documenting it here where there's discussion. — Gavia immer (talk) 03:14, 14 November 2010 (UTC)
Wow, that was fast. Thanks! PrimeHunter (talk) 03:16, 14 November 2010 (UTC)
Yeah, that was surprisingly fast. Good work, guys. D. F. Schmidt (talk) 20:20, 14 November 2010 (UTC)

Suggestions

1- Please create the Village pumb button to left section of website. Please this button to all wikimedia websites and all own languages. http://www.wikimedia.org/

Help About Wikipedia Community portal Recent changes Contact Wikipedia

Please add to bottom of these buttons in left section.

2- Please read: A personel appeal from Wikipedia founder Jimmy Wales.

There is this button in top of wikipedia. Please this button to all own wikimedia websites and all languages. http://www.wikimedia.org/

When I click to Please read: A personel appeal from Wikipedia founder Jimmy Wales., this page is open: http://wikimediafoundation.org/wiki/WMFJA1/en?country_code=TR

You must translate the this page to all other languages —Preceding unsigned comment added by 85.106.206.254 (talk) 09:44, 14 November 2010 (UTC)

It has been translated to nearly 40 languages so far. We're still working though.  :) Philippe Beaudette, Wikimedia Foundation (talk) 11:56, 14 November 2010 (UTC)
Once again, we have no control over what other language wikis do. If you wan to raise issues with them you should contact them directly. OrangeDog (τ • ε) 16:27, 14 November 2010 (UTC)
Philippe Beaudette, Wikimedia Foundation, how come when I click the Jimbo message in this English WP, I get a non-English message? I mean to ask, not the tech stuff, but the thoughts behind it. This is disturbing. -DePiep (talk) 20:28, 14 November 2010 (UTC)
It's grabbing your location via http://geoiplookup.wikimedia.org/ and setting the language via the country/area you're in, I believe. There are some issues with this because there are countries have "official" languages as well as one to several other widely used languages. I'm not sure if the fundraising staff is working on providing a "other languages" bar, but that might be helpful. Killiondude (talk) 22:22, 14 November 2010 (UTC)
Yes I could guess it worked like this, but the question stays: when linked from the English WP, why end up with a different language? Why does WP geotrack my IP at all? -DePiep (talk) 22:58, 14 November 2010 (UTC)
Geoip is used so it's known what portion of donations are coming from France, and make sure those donations go through the French Wikimedia Chapter (with some percentage going to the global Wikimedia Foundation, and some percent kept with the chapter). Same for Germany, the Netherlands, and other countries that have chapters and have fundraising agreements with the global Wikimedia Foundation. As for language, I think it's assumed if you are in France (or wherever), you speak French (whatever language). But perhaps it's not 100% true and we should think about making the banners/messages smarter in that regard. Cheers. --Aude (talk) 00:02, 15 November 2010 (UTC)
Nah. So no option for me. It's a "we determine what you are". -DePiep (talk) 00:09, 15 November 2010 (UTC)

Muhammad Cartoon DDOS Attack in 03/2010?

I was doing some anti-vandalism research when I happened to come across this: [4]

The Jyllands-Posten_Muhammad_cartoons_controversy article received 5.4 million hits on 3/9/2010.

This averages out to 62 hits/sec. for the entire day. Nearly 1500 hits/sec. were being recorded at peak (per [5]). Also unusual because nothing related to the controversy seemed to happen on this date?

Clearly there was some funny business. Anyone have any insight into what went on? Two plausible solutions in my mind: (1) statistics error, and (2) DDOS attack. If the latter, was there any discussion of it archived I could take a look at? West.andrew.g (talk) 20:38, 9 November 2010 (UTC)

I was going to say Everybody Draw Mohammed Day, but the dates don't quite match up. OrangeDog (τ • ε) 23:42, 9 November 2010 (UTC)
March 3, 2010 was the day that Final Fantasy XIII was (supposed to be) released. Was there a scene in that where you have to battle the Prophet? --Ludwigs2 00:01, 10 November 2010 (UTC)
Hmm, I've still got nothing. I do see that the folks who are upset about this kind of thing have launched plenty of DDOS attacks in the past, including against Wikipedia -- but I've found nothing definitive regarding this date. West.andrew.g (talk) 19:39, 10 November 2010 (UTC)
So someone launched a 1.5Tb DDOS attack against our servers... and we only noticed eight months later during a random statistics trawl? I'd call that a fail... ::D Happy‑melon 01:37, 15 November 2010 (UTC)
Yeah, 62 hits/sec is nothing. Could be classified as random noise. An appropriate measurement for a DDOS worthy of comment would be in GB's per second. Regards, SunCreator (talk) 02:03, 15 November 2010 (UTC)
On the morning of March 9, the Irish police announced the arrest of seven people accused of plotting to assassinate Swedish cartoonist Lars Vilks (who had drawn Muhammad as a pig in 2007, but not for Jyllands-Posten), generating considerable media coverage ( examples: [6][7]), but probably not enough to explain 5.4 million. (For comparison, the 115th anniversary of the discovery of x-rays, featured in a Google doodle last week, brought around 1.25 million additional pageviews[8].) And the page views of the article about Vilks himself reached only a maximum of 8.800 [9]. Regards, HaeB (talk) 02:45, 15 November 2010 (UTC)
Yeah, as far as DDOS attacks go this would be nothing impressive. However, 1500 hits a second is significant. It's highest rate (bucketed hourly) so far in all of 2010. In comparison, The Who has the second highest rate at around 154 hits a second -- and that peak occurred during the Super Bowl (where they were the halftime show). So while not an epic DDOS, there is some funny business.— Preceding unsigned comment added by West.andrew.g (talk • contribs) 4:08, 15 November 2010
I have written a short note about this for the Signpost (Wikipedia:Wikipedia Signpost/2010-11-15/News and notes#Briefly, to be published later today - corrections welcome). Regards, HaeB (talk) 07:50, 15 November 2010 (UTC)
Not enough page requests to count as a serious DDOS attack. Probably just a script kiddie. Offliner (talk) 14:02, 15 November 2010 (UTC)
With this going in the Signpost, I thought I'd leave a helpful note or two and some more interesting details. First, my statistical analysis only encompassed Jan. 1 2010 through Aug. 1 2010. Second, the top five or so traffic hours during this interval look something like:
* Jyllands-Posten_Muhammad_cartoons_controversy - 1474 hits/sec. - Mar. 9, 2010 - Reason unknown
* The_Who - 154 hits/sec - Feb. 7, 2010 -- Superbowl halftime show
* Main_Page - variable - This is where most instances (traffic hours) of the main page show up in terms of traffic
* Ice_hockey_at_the_2010_Winter_Olympics - 73 hits/sec. - Feb. 28, 2010 - US vs. Canada for the gold medal
* Dennis_Gabor - 55 hits/sec. - June 5, 2010 - He was the Google Doodle on that day
* Rebirth_(Lil_Wayne_album) - 51 hits/sec. - Feb. 2, 2010 - The album's release date
Thanks, West.andrew.g (talk) 18:35, 15 November 2010 (UTC)
Interesting, thanks! Is it 500 hits/sec or 154 hits/sec for The Who? Regards, HaeB (talk) 19:13, 15 November 2010 (UTC)
Apologies, 154 hits/sec. is correct (and now changed above). West.andrew.g (talk) 19:19, 15 November 2010 (UTC)

Why does this page appear to exist?

Why does MediaWiki:Common.css/Foo appear to exist, yet have no history [10] and no meaningful content? Should an admin WP:G6 it? Can an admin do so? Why does this work for any value of "Foo"? --NYKevin @924, i.e. 21:10, 13 November 2010 (UTC)

That default values gets show on every (non existing) subpage of Commons.css, eg. MediaWiki:Common.css/thisisanuniquevaluenobodywouldevercreate - Hoo man (talk) 21:31, 13 November 2010 (UTC)
If it wasn't automatically admin restricted then it would probably be created soon after you made such a tempting link! Have you seen the deletion log [11] for The weather in London which used to be an example of what not to create? PrimeHunter (talk) 22:54, 13 November 2010 (UTC)
I know that you can only edit interface message with the editinterface right, don't worry :) - Hoo man (talk) 00:13, 14 November 2010 (UTC)
This still leaves the question of why it "exists" at all. Why not just leave it a red link? It's already impossible for non-admins to edit Mediawiki: space pages anyway, what does this accomplish? --NYKevin @304, i.e. 06:17, 15 November 2010 (UTC)
Subpages of existing interface pages are used for translating into different languages. The /Foo subpage allows you to localize the message into the Foo language. If the message has not been localized into the Foo language, it falls back to English. That's what is happening in this case. The English version is shown because there's no translation for the Foo language. Reach Out to the Truth 20:18, 15 November 2010 (UTC)

rationale-question

What's the rationale for new users' having to wait 4 days/10edits before they can upload files? Any link to past discussions/decisions? Choyoołʼįįhí:Seb az86556 > haneʼ 17:12, 15 November 2010 (UTC)

Assuming we're talking about straight uploads rather than upload-over-another-image, I presume it was was thought that adding "File: Moose grazing.jpg" to an article whilst simultaneously uploading said file as a shock image was too tempting a possibility for the casual vandal. 10 edits and 4 days is a very low bar anyways, and the "confirmed" right can now be granted to overcome the time impediment. Couldn't point you to anything though, sorry. - Jarry1250 [Who? Discuss.] 18:13, 15 November 2010 (UTC)
  • Those limits are intended to thwart automated text-spam accounts, and so a person could just edit 10 random pages (such as for simple clarity of wording) to get beyond the limit. The related limits prevent any would-be WP:Sockpuppet acounts from editing a protected article as if being a totally new user, during the first few days. -Wikid77 (talk) 18:23, 15 November 2010 (UTC)
thanks. Choyoołʼįįhí:Seb az86556 > haneʼ 04:15, 16 November 2010 (UTC)
This Google search leads to this Signpost article, which mentiones that uploads were restricted to autoconfirmed users because of bug 12556. Graham87 10:23, 16 November 2010 (UTC)

Pending Changes feature update

We are currently planning to roll out a new version of the FlaggedRevs extension to all wikis on WednesdayTuesday, November 23 starting roughly 3:15pm PST (23:15 UTC). This is used for Pending Changes on en.wikipedia.org and Flagged Revisions on many other wikis. This will have a new reject button, some diff page load optimizations to help complicated diff pages load faster by displaying the diff prior to displaying the old revision, and many under-the-hood code improvements. We decided to delay the rollout until next week (previously scheduled for this week). It took us a little longer than we liked getting the exact configuration of FlaggedRevs we plan to test, and we wanted to make sure people had more time to play with this before we went out with it.

We have several test environments in place with FlaggedRevs/Pending Changes configured:

Please let us know if you have any problems. Thanks! -- RobLa-WMF (talk) 06:14, 16 November 2010 (UTC)

Oops, fixed thinko - meant to say "Tuesday" -- RobLa-WMF (talk) 07:20, 16 November 2010 (UTC)

webcite down

Does anbody know somethink about the outage of http://www.webcitation.org ? I just wanted to let my bot check some urls and got an UnknownHostException. ICMP ping is ok, but no TCP connection is accepted. Merlissimo 16:43, 16 November 2010 (UTC)

I just checked the site, and everything appears to work fine now. Give it another try. — Huntster (t @ c) 17:10, 16 November 2010 (UTC)
Yes, website is back. Merlissimo 17:21, 16 November 2010 (UTC)

Blocking at Commons

I have been blocked from contributing to Commons with the following message You do not have permission to edit this page, for the following reason: Your IP address, 41.213.126.5, has been automatically identified as a tor exit node. Editing through tor is blocked to prevent abuse. Can anyone help? Androstachys (talk) 09:31, 14 November 2010 (UTC)

Editing through Tor isn't generally allowed. If you're not actually using Tor, you can have someone (like me) request an account on your behalf at commons:Commons:Administrators' noticeboard (once you have an account, you can log in and edit normally). Just specify a username and explain your situation in more detail. Dcoetzee 09:55, 14 November 2010 (UTC)
I don't have the faintest idea what Tor is...... but have been uploading images since 5 January 2010 and been using the same service provider all that time - nothing has changed, which is why this is a puzzle. The same thing happened to me some months back on WP and some kind soul got me an exemption. I suppose 'Androstachystemp' would be a suitable username if I don't want accusations of sockpuppetry. Thank you for the offer. Androstachys (talk) 13:24, 14 November 2010 (UTC)
Tor is a service that masks your real IP address. If you're editing through a proxy server and someone else behind the same proxy is using Tor, then you may get this message. --NYKevin @306, i.e. 06:20, 15 November 2010 (UTC)
I see there is a Tor blocking bug that has just been fixed. Doesn't seem related but there were some changes in Tor blocking about the time of your problem. Does your problem still exist? TOR nodes not being properly blocked. Darrell_Greenwood (talk) 02:21, 16 November 2010 (UTC)
Bear in mid that Tor exit nodes can be dynamically allocated IPs or indeed further proxied. Rich Farmbrough, 16:27, 17 November 2010 (UTC).

Raising the MediaWiki expression-nesting limit

15-Nov-2010: We are wondering when the MediaWiki parser nesting limit will be set higher, to allow the nesting of if-else logic, template transclusions, and parameter values to exceed 40-levels deep (such as allow 100 deep). The following simple example from Plastikspork, shows the problem for invoking 41 nested templates, which exceeds the "expansion limit": 

Nest 41 templates: {{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|
{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|
{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|{{1x|
{{1x|{{1x|{{1x|{{1x|{{1x|
{{1x|foobar-data}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}
}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}}

Nest 41 templates:


foobar-data

The reddish error message, displayed as "{{Expansion depth limit exceeded|foobar-data}}", is triggered when the nested logic exceeds the parser limit.

I did a "search" of the technical Village Pump, but only a few pages matched "expansion depth" about the expression-nesting limit. As a computer scientist, I would have expected a limit of at least 100 deep, perhaps 500, or unlimited nesting if-else clauses or #expr expressions. The measurement converter, Template:Convert needs the limit raised, beyond 50 nested levels, to expand the template to "super-clever" coding where it can spot common-sense problems, using heuristic tests of several nested if-else expressions. Currently, we are having to make "spoon-fed" changes to templates, as if working on some dinky, primitive computer system. Someone please let the developers know to set the "limit=60" or "limit=500" as soon as possible. This is a classic issue of where WP:PERF does not mean to allow Wikipedia to perform like an archaic computer system which can't handle more than 40 validation checks. This has been a problem for over 2 years, and we had assumed common sense would have fixed it by now.

As a work-around, some if-logic can be coded as thousands of non-nested if-expressions (not nested as if-else-if-else-if-else...), or limit the "else" clauses to within 40 levels. Also, when the if-logic all checks one item, then a #switch can be used with hundreds of branches. However, a so-called "guarded command structure" (used over 35 years in computer science) cannot be nested beyond 40 conditions, to pre-scan various data items for validation checks, before proceeding further in the template coding. When templates are used inside other templates, they each subtract from the total of 40 levels, leading to cases where even a 4-deep coding of if-else will die.

More about this, later. -Wikid77 (talk) 18:23, 15 November 2010 (UTC)

See WP:BUGZILLA. --NYKevin @834, i.e. 19:01, 15 November 2010 (UTC)
I think that any developer will tell you that where you need a 40 level deep template transclusion; "you are doing it wrong". In such cases, most likely what is actually needed, is someone developing a PHP MediaWiki extension, to solve the problem properly. Unfortunately extensions don't write themselves. This is the current set performance limit, and I doubt it will change, but indeed you can always try on bugzilla. —TheDJ (talk • contribs) 20:48, 15 November 2010 (UTC)
  • Well, that would explain why common sense hasn't worked to increase the limit, after 2 years. I guess they would advocate hammers with no claws ("if you need to remove nails, you're hammering wrong" or similar crap). It's like those do-it-yourself TV shows, where they actually hit a nail 23 times to bend in a knot-hole, then remove that from the film, with a re-take of hitting another nail 4 times in soft wood. Anyway, thanks, I'll search on bugzilla. -Wikid77 (talk) 21:50, 15 November 2010 (UTC)
Common sense would suggest you should learn how to use a hammer properly before attempting any DIY. OrangeDog (τ • ε) 22:28, 15 November 2010 (UTC)
I agree that increasing the limit would help. What would help even more would be some more advanced parserfunctions which would reduce the complexity of templates like {{convert}}. Like say "precision" to determine the number of significant figures, or even better, something that does what {{Rnd}} does. Of course complaining here may be "preaching to the choir". Plastikspork ―Œ(talk) 01:59, 16 November 2010 (UTC)
If someone wants to write a parser function for precision, it will be easy: look for a decimal point, then count digits to the end or "E" (scientific notation: 1.72E+9), else the precision=0 if no decimal point. -Wikid77 (talk) 12:42, 16 November 2010 (UTC)
Maybe just exposing a printf workalike would solve a wide range of formatting problems. Wondering if {{Rnd}} could be recoded so it does not need so many sub-templates, which are just a workaround of the days before parser functions.--Salix (talk): 07:11, 16 November 2010 (UTC)
  • I have recoded for a shorter version of {Rnd}, and that improvement is underway. Also, although more parser functions would help, a little, the reality seems to be that a nesting limit of 40 is just too miserly, to handle all the options needed for a real-world application. I have calculated Wikipedia to reach 9 million articles, and it is easy to see Commons will gain 50 million images (pages from old illustrated books). Most of Convert was first implemented when MediaWiki was even more primitive, so we know how to compensate for the restricted performance allowed by the MediaWiki software. Just think of this as the Dark Ages of Wikipedia, waiting for a Renaissance of better technology as in older computer systems (which could nest 300 if-else clauses). Computers were formerly very expensive, and mistakes then were costly, so a limit like 40 would have been fixed, quickly, as a tremendous waste of valuable computer time and man-hours. Anyway, for the present, the very efficient Template:Precision can be reduced by 1 nest-level (from 4 templates to 3) by combining the contents of Template:Precision/a, where parameter {2} would be expanded as a Titleparts function:
         {{#if:{{{2|}}}|{{#expr:(ln{{{2}}})/ln10}}|{{#expr:{{precision/{{#expr:3*{{{1}}}>{{{1}}}0}}|{{{1}}}}}}}}}
    That expanded template coding, replacing the two instances of {2}, would become the new contents of {Precision}, and we could bypass template {Precision/a} as 1 less level of nesting. I dread that people (especially volunteers) have to penny-pitch resources this way, but it takes time for a bureaucracy to make improvements, even to fix the most crucial of underlying resource limitations. Meanwhile, if someone wants to write a parser function for precision, the idea is simple: look for a decimal point, then count digits to the end or "E" (scientific notation: 2.67E+13), else the precision=0 when no decimal point. Thanks to all for considering these issues. -Wikid77 (talk) 12:42, 16 November 2010 (UTC)
  • Simple string parser functions are available as an extension. Instead of releasing these we are told we don't need them and end up using incredibly inefficient string handling templates that lead to ridiculous limits ("this template won't work if the category name is more than 50 characters or contains UTF8") and even results in unscalablility (template transclusion limit exceeded). Rich Farmbrough, 16:33, 17 November 2010 (UTC).
The appropriate place to ask for StringFunctions (or indeed, string functions) is [12]. Unfortunately, it's already marked WONTFIX. Also, of any WONTFIX bug, SF has the most votes [13], so voting probably won't do anything (OTOH the dev's don't like it when people comment on bugs unless said comments are useful). Maybe we should ask WMF to enable SF (since the dev's don't seem very interested in doing so)? --NYKevin @745, i.e. 16:52, 17 November 2010 (UTC)
I had the last comment on the bug in September 2009. I have reopened the bug. Rich Farmbrough, 17:42, 17 November 2010 (UTC).

ENGVAR editnotices

A discussion at WP:VPL made me wonder about the feasibility of using similar system for ENGVAR editnotices as we currently have for the BLP editnotice. The latter loads via MediaWiki:Common.js based on the mere existence of Category:Living people in an article; in principle we could do something similar with Category:British English articles etc. This would be better than the current system of eg {{British-English-editnotice}}, as seen at Amnesty International, applied via the appropriate editnotice page (Template:Editnotices/Page/Amnesty International), since this needs admin permission and as a result isn't too widely used. So my question is - is there any technical reason why this would be a problem? Rd232 talk 09:45, 16 November 2010 (UTC)

Category:Use British English is the extant category. Rich Farmbrough, 17:31, 17 November 2010 (UTC).

Accessing better string parser-functions

I've noticed http://www.mediawiki.org/wiki/Extension:StringFunctions describes some basic string functions for character-handling work. Many of those functions should be made standard for everyone, and I saw a related page for MediaWiki 1.7. As several users have noted, the writing of templates would be much easier if those simple string functions were always available, rather than using "expensive" nest-limited templates (such as Template:Str_len), due to the current MediaWiki setup which severely limits template and if-else nesting to a mere 40 levels.

We should have many of those string-functions as standard, such {{#len:xxxxx}} to get string length, or {{#explode:xxx.com|.|1}} to extract "com" or other string parts. Using such functions, a simple way to get numeric precision would be: {{#len:{{#explode:45.620|.|1}}}}, to get 3 as the precision for number 45.620. Currently, use of {{#len:xyz}} returns the result: {{#len:xyz}}, where a result of "3" is the length of "xyz".

The severely minimal limits for Wikipedia's template work are discussed (above) in "#Raising the MediaWiki expression-nesting limit". As in any system, a well-chosen set of basic support functions can replace a larger peculiar set of weird functions which were formerly preverted to try to do simple things with a spastic mixture of unusual functions. Wikipedia's limited functionality has been offset by just such a collection of contorted mixtures, trying to do the simplest of operations by twisted use of bizarre parser functions, such as within Template:Substring or Template:Str_sub. Wikipedia's intelligent users have shown that they cannot be stopped by primitive software, but it would be much easier to just provide the basic support functions they need, directly.

Has anyone set a date as to when {# len:xxxx} or {#explode:xx.x|.|1} will become standard parser functions for all users, or what is the planned date for upgrade to version MediaWiki 1.7? -Wikid77 (talk) 15:38, 16 November 2010 (UTC)

The date is apparently "never". StringFunctions has been finished for years, but was never activated because the devs feared they might overload the servers. Instead, they opted for another method of string manipulations based on yet another language. You can find the entire discussion in Bugzilla. — Edokter • Talk • 15:46, 16 November 2010 (UTC)
Performance was originally the problem, but that was solved and eventually merged into ParserFunctions. But the functions are not enabled because they would prefer to embed a real programming language such as Lua. But they are also not interested in working on embedding a real programming language right now, and any embedded programming language must be able to run securely without PHP extensions or external programs so people running MediaWiki on really crappy webhosting can still copy our templates. This old thread has a humorous summary of the years-long discussion regarding StringFunctions. Anomie⚔ 19:02, 16 November 2010 (UTC)
  • Thanks for that background info: sounds like a typical "paralysis of analysis" in hoping how future changes would lead to a better solution if nothing is done now. In reality, we have Template:Str_len (Feb. 2009, used now in more than 30,000 articles), as a clever Rube Goldberg machine to find the string length using a complex algorithm, which {{#len:xxxx}} could have performed by a single quick parser-function call. I would guess {Str_len} is 500x times slower, maybe 2,000x slower? However, I can imagine how other string parser-functions might have been the larger worry, among the total StringFunctions. In such cases, just compromise by adding a few string functions to the standard set. Meanwhile, I might be removing some uses of {Str_len}, in some templates, where it is not absolutely needed. -Wikid77 (talk) 06:34, 17 November 2010 (UTC)

I really think this is something that ought to be sorted - it's idiotic to have extremely inefficient hacks being used when there's a better solution ready and waiting to be switched on. Does anyone have the ear of someone with the power to make such a decision? (As long we don't provoke the absurd opposite reaction of disabling the current hacks, which is what I believe some in developer-world would like to do.)--Kotniski (talk) 07:18, 17 November 2010 (UTC)

It's interesting to see how many of the current threads on VP(T) are affected by this strange situation. Rich Farmbrough, 17:34, 17 November 2010 (UTC).

Now that I think about it, this is a perfect time to ask for StringFunctions... I pledge to donate $10, if StringFunctions are enabled. There... — Edokter • Talk • 01:01, 18 November 2010 (UTC)

Problem with the radio buttons for donating to Wikepidia

Click on Read Now ....for "Please read: A personal appeal from Wikipedia founder Jimmy Wales". When you get to the next screen, click one of the dollar amount radio buttons to donate and then press the Paypal button. NOTHING happens. I was able to donate by using the Donate button on the LHS of the Wikipedia screen. There seems to be a bug in the system not allowing donations from that page. It's asif the page is totally NON interactive. —Preceding unsigned comment added by 70.250.187.102 (talk) 15:53, 17 November 2010 (UTC)

Unable to reproduce. Problems: "What browser is the user using, what is LHS ?" —TheDJ (talk • contribs) 23:45, 17 November 2010 (UTC)
"LHS" means "left-hand side", presumably. Graham87 02:14, 18 November 2010 (UTC)

italic article title

Is my browser broken, or why is Manufacturing Consent: The Political Economy of the Mass Media's title unitalicized despite the use of {{italictitle}}? --Cybercobra (talk) 07:34, 16 November 2010 (UTC)

Check the template's docs. :-) Killiondude (talk) 07:55, 16 November 2010 (UTC)
Yep, too long for the template to handle, I think (and that's because of the devs' bizarre stubbornness in refusing to allow efficient string handling in templates). I've changed it to an explicit DISPLAYTITLE, and it seems to work now.--Kotniski (talk) 10:04, 16 November 2010 (UTC)
Note there is a long version of some (one?) of the string handling functions, if that helps. Rich Farmbrough, 17:26, 17 November 2010 (UTC).
Where is the consensus for adding {{italictitle}} or an explicit italics DISPLAYTITLE to articles like this? Thanks. – ukexpat (talk) 18:10, 18 November 2010 (UTC)
There was a long discussion on Wikipedia talk:Article titles. There seemed a majority in favour some form of italics in titles.--Salix (talk): 21:50, 18 November 2010 (UTC)

Sitenotice renders after page load, causing annoying jump

  Unresolved

I don't have a problem with the big fundraising banner per se, and thanks for giving me the option to dismiss it, but when pages load it appears that everything renders and then the image loads a fraction of a second later, causing for me at least (Firefox 3.5, Linux) all the content beneath it to jump down by 170px or so. More than once since the present fundraising drive began I've inadvertently clicked on it when I've aimed for a link in the article text and then the banner has loaded under my mouse pointer. I'm familiar with this problem for static HTML pages (most commonly encountered with long photo galleries where it takes a long time to load all the pics), and would usually recommend sorting it out by specifying the width and height properties in the the img tag. This one however seems to involve javascript, so perhaps it's not quite so straightforward to fix. Anyway, I and many others would greatly appreciate it if someone could fix this such that a placeholder of the correct dimensions is included when the page renders and then the image loads into it without affecting the layout. I'm sure this is not beyond the capabilities of the collective wisdom of Wikipedia's technical community. Thanks, a reader. 82.32.184.4 (talk) 20:00, 16 November 2010 (UTC)

Yes, this is discussed earlier on the page (#Links to sections). I agree that it's annoying and that's mostly why I closed the fundraising banner from my view. Killiondude (talk) 20:21, 16 November 2010 (UTC)
So the problem has been noted, but there are no plans to do anything about it? 82.32.184.4 (talk) 11:35, 18 November 2010 (UTC)

Database dumps

Anyone know the latest status on these? Rich Farmbrough, 17:32, 17 November 2010 (UTC).

11-10-2010 - New errors at Wikitech. – Allen for IPv6 17:49, 18 November 2010 (UTC)

Category section testing

Take a look at Category:Unreferenced_BLPs_from_October_2008. Is there anyway that the letter section can be worked out so that it can be automatically summarized the parts that are empty or still have items remaining as follows:

  • A - B - C - D - E - F - G - H - I - J - K -L - M - N - O - P - Q - R - S - T - U - V - W - X - Y - Z

The sections that are empty may have changed since above was posted, but you'll get the idea. Regards, SunCreator (talk) 02:43, 16 November 2010 (UTC)

There is no practical way as far as I know. The letter section is made by {{Category TOC}} which doesn't have access to that information. {{PAGESINCATEGORY}} at mw:Help:Magic words#Statistics only works for a whole category, and it's expensive. It might be possible to categorise each article in both Category:Unreferenced BLPs from October 2008 and a category specific to the initial letter, and then a variation of {{Category TOC}} could test whether the letter category is empty, but I wouldn't recommend this. And I don't know how many expensive {{PAGESINCATEGORY}} would be possible. PrimeHunter (talk) 04:06, 16 November 2010 (UTC)


I or another programmer could write a tool for this, but obviously it would be a separate webpage, potentially slow to load and not integrated into the main site. Would that be useful?
The only on-wiki way to do this would be to sub-cat them by first letter, I think. Hm, yes that's not hard. Rich Farmbrough, 16:34, 17 November 2010 (UTC).
Except that being people, the first letter of the article is not necessarily the letter it is sorted by. So easy becomes very tricky. Otherwise, piece of cake (in fact I did it I just need to revert self now). Rich Farmbrough, 17:23, 17 November 2010 (UTC).
(If anyone can bridge that gap, let me know.) Rich Farmbrough, 17:24, 17 November 2010 (UTC).
Thanks for the replies everyone. A simplified variation requiring parameters instead of automatically working out the categories has been provided with: Template:URBLPRLetterProgress and is now being used. Regards, SunCreator (talk) 21:57, 19 November 2010 (UTC)

This is a minor edit (what's this?); does not open in new window and I lose my work

On the computer I am on today, there is an Ajax list of edit summaries below "Edit summary", under the box where I type. I clicked on the correct choice and somehow ended up clicking on "What's this?" (above "Show preview") and lost my work. Fortunately, it wasn't a lot. Perhaps "what's this?" shouldn't send me to a new page but should let me continue editing on the same page, like "Editing help".Vchimpanzee · talk · contributions · 16:51, 18 November 2010 (UTC)

MediaWiki:Minoredit is where that message lives, but it seems that message will not accept HTML so it cannot easily be made to open in a new window. Personally, I just add span#minoredit_helplink { display:none; } to my skin.css, which hides it completely (I do the same for much of the other clutter down there). Anomie⚔ 17:36, 18 November 2010 (UTC)
Well, just thought I'd ask. Thanks. It never or rarely happened before.Vchimpanzee · talk · contributions · 21:14, 18 November 2010 (UTC)
I noticed the other day that the "Edit summary" link does the same when I bumped it going to hit "Save page". The workaround is nice, but usurping the edit window by default isn't a great approach for new users. —Ost (talk) 22:41, 18 November 2010 (UTC)
I wouldn't worry about new users, won't they have "Warn me when I leave an edit page with unsaved changes (Vector skin only)" checked automatically? - Kingpin13 (talk) 17:10, 19 November 2010 (UTC)
I was having this problem with IE a couple of years back; someone coded up a script for me that you could probably adapt for your browser/skin. 「ダイノガイ千？！」? · Talk⇒Dinoguy1000 19:55, 19 November 2010 (UTC)

Opening external links in a new window

  Resolved

Can any technically minded bod devise some code that I can add to my .css or .js file that will automatically open external links in a new window? I know the option exists as a gadget here on enwiki, but I'd like to have the functionality on other projects too. —  Tivedshambo  (t/c) 16:18, 20 November 2010 (UTC)

You could copy it from here: MediaWiki:Gadget-exlinks.js. ―cobaltcigs 17:07, 20 November 2010 (UTC)
Perfect - thanks. —  Tivedshambo  (t/c) 17:43, 20 November 2010 (UTC)

SVG Text

Could someone correct the text in the 3 svg illustrations at commons:Category:Federal Justice Statistics 2008–Statistical Tables? Wikipedia:SVG Help seems dead.Smallman12q (talk) 18:31, 20 November 2010 (UTC)

You can ask at commons:Commons:Graphic Lab. --NYKevin @954, i.e. 21:54, 20 November 2010 (UTC)

Project page title problems at Star Trek WikiProject

Hi, the WP:WikiProject_Star_Trek project page is displaying "Special:RecentChangesLinked/Wikipedia:Version 1.0 Editorial Team/Star Trek articles by quality log" as its title in IE8, instead of "Wikipedia:WikiProject Star Trek". I've attempted to fix this problem, but keep getting reverted, see: [14] and [15]. Is there a way to display the recent changes for the project's articles, as is done by the toolserver here: Wikiproject Watchlist - WikiProject Star Trek, and transclude the RCs onto the main project page w/o messing up the title? Thanks in advance, --Funandtrvl (talk) 20:15, 20 November 2010 (UTC)

I tried, but I could not fix it. This should go to Bugzilla. — Edokter • Talk • 20:47, 20 November 2010 (UTC)
I don't have a bugzilla link, but this is a known bug when transcluding the recent changes feeds. Quite honestly, it's surprising enough that transcluding recent changes works at all; the bug is pretty minor compared to that. — Gavia immer (talk) 21:17, 20 November 2010 (UTC)

Update on the Article Feedback Tool

Hi all! On behalf of the Article Feedback Tool workgroup, I would like to announce phase two of the tool pilot. Here's a little background first, for those of you who aren't familiar with the Article Feedback Tool:

  • The AFT is used for article rating and is used on articles that are part of WikiProject United States Public Policy. A full list of pages that have the tool enabled may be seen at Category:Article Feedback Pilot.
  • For an example of what the AFT looks like, please see any page in the above category, such as 5 centimeter band.
  • The tool was launched on 22 September and has since seen over 12,500 ratings submitted since that time. This page details the data that has been gathered up to 8 November.
  • The FAQ contains more information.

So far, the ratings of only articles in the USPP WikiProject are being compared to their revision histories. Phase two of the pilot involves the AFT being used on a small number (about 50) of non-USPP articles in an effort to better analyze the value of these ratings. These articles are being listed at mw:Article feedback/Public Policy Pilot/Additional Pages (please help by adding suggestions there), and are expected to undergo significant edits in the near future—such as upcoming elections and films. See also mw:Article feedback/Public Policy Pilot/Design Phase 2 and mw:Article feedback/Public Policy Pilot Phase 2.

Everyone's help is appreciated in taking a few seconds to rate an article! Please leave any feedback about the tool at this page. Please try to keep questions below this thread to avoid fragmented discussion, but User:Howief should be able to answer any private queries. Thank you! /ƒETCHCOMMS/ 01:44, 21 November 2010 (UTC)

Pywikipediabot error

I get an error message with a Python Pywikipediabot script: meta:Uploadmultiple.py. All of the other scripts including meta:Upload.py work but not this script.

I was attempting to upload two images and here is the error message: 

C:\Users\t\Desktop\pywikipedia>Uploadmultiple.py
unicode test: triggers problem #3081100
Traceback (most recent call last):
  File "C:\Users\t\Desktop\pywikipedia\Uploadmultiple.py", line 62, in <module>
    import wikipedia, config, mediawiki_messages
ImportError: No module named mediawiki_messages

pywikipedia.py/mediawiki_messages.py "Allows access to the MediaWiki messages, that's the label texts of the MediaWiki software in the current language"

Unfortunately there is not a big support structure for pywikipediabot. I posted my question on the IRC channel and emailed the creator of this script with no response. Thank you so much in advance! Adamtheclown (talk) 18:31, 19 November 2010 (UTC)

I gave up, too complex for me to understand and no one anywhere has volunteered to help. I am now trying to run commonist, which is presenting its own problems. * sigh * Adamtheclown (talk) 21:03, 21 November 2010 (UTC)

Talk page size issue

I'm pretty sure that my talk page is not the fifth largest talkpage among all project spaces, despite what Wikipedia:Database reports/Talk pages by size claims. Is this a DB issue or an issue with the bot? Veinor (talk to me) 21:58, 20 November 2010 (UTC)

The bot says it counts all the subpages so I guess it is counting all this which includes link count pages. Rambo's Revenge (talk) 22:05, 20 November 2010 (UTC)
(edit conflict)It's counting everything listed in Special:Prefixindex/User talk:Veinor, not only your talk page archives, although I can't tell if it's counting page history or only the last revision. Really, though, this just makes the more general point that I already raised due to the bot spam on Talk:Main Page: These notifications are useless nonsense, except where they are harmful nonsense. I don't know how anyone thought this was a good idea. — Gavia immer (talk) 22:09, 20 November 2010 (UTC)
Well, they reminded me that the link count pages exist and should probably be purged as useless, so they weren't completely useless! Veinor (talk to me) 22:11, 20 November 2010 (UTC)
What is the possible purpose of this report? And what is the message at Wikipedia talk:Reference desk to be more efficient in the use of kilobytes? Archiving is archiving, it's a page that has been around almost since the start of the project. Franamax (talk) 23:41, 20 November 2010 (UTC)
I think the primary purpose is to satisfy some people's curiosity (including mine). And I would ignore that massage, there is no problem with large archives. Svick (talk) 13:27, 21 November 2010 (UTC)

Another secure server/interwiki issue

Issue: I am logged in to the same Firefox session as both User:Jclemens and User:Jclemens-public. Both are global logins. When I click interwiki links from this (Jclemens, administrator, secure server-using) account, I end up on the other project site as Jclemens-public. I've seen and adopted a monobook.js script that looked like it would fix this for external links, but I've not seen any improvement. Anyone else had this and/or know how to fix it? Jclemens (talk) 02:05, 21 November 2010 (UTC)

There is a script installed on English Wikipedia and Mediawiki Commons for ALL users now that rewrites such links these days. Perhaps your own script is interferring with the default script, preventing proper execution. —TheDJ (talk • contribs) 12:53, 21 November 2010 (UTC)
Problem with that is that I didn't have any script installed at all previously, and I was still having this problem. It also happens when I type "m:foo" or "wikt:foo" into the search bar. I can take all scripts off and see, but it's not like I had anything going on before when I first noticed this... Jclemens (talk) 16:21, 21 November 2010 (UTC)
Well that should no longer happen (for the past months). The only thing that cannot be fixed is the addressbar, because that is a hardwired redirect in the server configuration. —TheDJ (talk • contribs) 17:15, 21 November 2010 (UTC)
By addressbar, do you mean the browser address bar, or the "search" bar on the Monobook skin? If it's the former, I completely understand, but the latter seems a bit disappointing. I've confirmed that if I follow a meta: link it comes across as secure, but if I enter "Commons:User:Jclemens" then it dumps me back into the insecure mode. Jclemens (talk) 00:19, 22 November 2010 (UTC)
I mentioned this back in July at Wikipedia:Village_pump_(technical)/Archive_78#Glitch_with_script.Smallman12q (talk) 01:22, 22 November 2010 (UTC)
Thanks for that. I'd looked, but there were a lot of hits for those terms in combination. Jclemens (talk) 01:28, 22 November 2010 (UTC)

Edit on arrival

I know I brought this up before, but I would like a second go on community views on a new feature to "edit on arrival"; immediately enter the edit window upon clicking any internal link. Meta has it; I wonder how difficult is it to implement it here too. Rehman 04:19, 21 November 2010 (UTC)

Seems terribly confusing to me in daily usage. Why don't you try using something like Navigation popups instead ? —TheDJ (talk • contribs) 12:51, 21 November 2010 (UTC)
Popups require clicking/pausing over the links, which takes time. It also doesn't work well on crappy browsers like IE, when you got no choice. Is it hard to add the option in preferences, like Meta did? Rehman 14:25, 21 November 2010 (UTC)
Of course it's not hard, but stuff there does need to be used by a substantial amount of people. Otherwise the list will become gigantic and hard to understand. For smaller utils, people best use their own script file. —TheDJ (talk • contribs) 17:13, 21 November 2010 (UTC)

A saner gadget would prepend “(edit)” links to each row of category content listings and to certain special pages (whatlinkshere, recentchangeslinked, and a few others). ―cobaltcigs 20:22, 21 November 2010 (UTC)

This would be best done as a user script because most users would be horribly confused and annoyed at a feature that doesn't let them read an article upon clicking a link. /ƒETCHCOMMS/ 21:13, 21 November 2010 (UTC)

Ok. So if I have to do it as a user script, how exactly do I do it? (I have no clue of these user script stuff) Rehman 00:23, 22 November 2010 (UTC)

Intermittent 403 Forbidden errors when returning to articles using the back button

Four times in the past three months (the fourth time was earlier today), I've been on a Wikimedia site (Wikipedia the first two times and the fourth time; Commons the third time), gone to a non-Wikimedia site using the same browser tab, and then clicked "back" to return to Wikimedia, only to be confronted by an error page saying:

Error 403 Requested target domain not allowed.

Requested target domain not allowed.

Guru Meditation:

XID: 896226428

Varnish cache server

(That number changes each time.)

The error page does not go away on a refresh, nor on access to other pages on the site; it takes a while before I can use the site again.

The first two times, I assumed the site was down; the third time, I got suspicious, partly because of the lack of comments on ezyang's Wikipedia Status page the previous time, and partly because all three times it had been fine before going to the external site. So I tried viewing the Recent Changes through a proxy site, and... I was able to access it that way, even though accessing Commons directly was still failing. (I also ran a tracert to Commons, which worked fine.) The proxy worked the fourth time as well.

Presumably, something is going awry in filtering bad HTTP requests, resulting in perfectly legitimate accesses being denied.

A few more potentially relevant data points:

  • I was using Internet Explorer 7 during all four incidents. (This has happened both on XP and Vista.)
  • I wasn't logged in during any of those incidents (but I'm not usually logged in).
  • The only program I have on my system I can think of that could be altering my HTTP requests is Norton 360, but as far as I know it's set up not to do that.
  • On the second and third incidents, the external page was the same (http://www.oregonbirds.org) but that might just be coincidence as those two occurred close together. (I don't remember what page it was the first time, but it may have been http://www.spaceref.com/news/viewsr.html?pid=34637, because I remember that the Wikipedia article I was trying to go back to was Kepler Mission, and I remember reading that linked page. The fourth time it was http://www.lscheffer.com/malbolge.shtml, which is linked from the Malbolge article.)
  • My ISP is Comcast.
  • In every case, I was on a perfectly ordinary Wikipedia/Commons page of the form http://en.wikipedia.org/wiki/[page title] (or commons.wikimedia), not something that one would expect to be 403 Forbidden.
  • I have never seen this specific message on any non-Wikimedia site, and those four are the only times I've seen it on a Wikimedia site.

Lucky Wizard (talk) 02:37, 18 November 2010 (UTC)

Anyone looking at this? This just happened to me again -- and this time it didn't fit the profile I sketched above.
I was on http://en.wikipedia.org/w/index.php?title=Talk:Main_Page&diff=397962046&oldid=397956117, and clicked "back" to return to http://en.wikipedia.org/w/index.php?title=Talk:Main_Page&action=history, only to get the message. This is the first time it's happened when clicking "back" from somewhere on a Wikimedia site; the other four times I got this message, I was on an external page before I clicked "back".
Incidentally, when I have the problem, I can still access other Wikimedia sites like de.wikipedia.
Lucky Wizard (talk) 04:14, 21 November 2010 (UTC)
It's almost certainly a problem with your ISP rather than Wikipedia. T. Canens (talk) 16:47, 22 November 2010 (UTC)

Rendering problems in Firefox

Several pages rendes wrong in my version of firefox(version 3.6.12). The only plugin that shows in my list is F-secure Browsing protection Toolbar. For example see this screenshot File:Screenshot-WP-19-11-2010.jpg of Coliseum at Richfield. Taemyr (talk) 09:31, 19 November 2010 (UTC)

Loading the page gives the following warnings, in norwegian:

Advarsel: Ukjent pseudo-klasse eller pseudo-element '-webkit-input-placeholder'. Regelsettet ignorert på grunn av ugyldig selektor. Kildefil: http://bits.wikimedia.org/skins-1.5/vector/main-ltr.css?283-12 Linje: 368

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 18

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 74

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 76

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 282

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 283

Advarsel: Ukjent egenskap 'zoom'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 285

Advarsel: Klarte ikke tolke verdien for 'filter'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 347

Advarsel: Ukjent egenskap 'zoom'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 360

Advarsel: Ukjent egenskap 'zoom'. Deklarasjonen ble sløyfet. Kildefil: http://bits.wikimedia.org/w/extensions/UsabilityInitiative/css/vector/jquery-ui-1.7.2.css?1.7.2y Linje: 398

Advarsel: Ukjent pseudo-klasse eller pseudo-element 'first'. Overflødig kombinator. Kildefil: http://en.wikipedia.org/wiki/Coliseum_at_Richfield Linje: 0

Advarsel: Klarte ikke tolke verdien for 'background-position'. Deklarasjonen ble sløyfet. Kildefil: http://en.wikipedia.org/wiki/Coliseum_at_Richfield Linje: 12 —Preceding unsigned comment added by Taemyr (talk • contribs) 09:39, 19 November 2010 (UTC)

These warning are really just messages that Firefox produces when it comes across css that it doesn't understand. In this case mostly css designed for other browsers. I have the same version of Firefox (albeit on a MAC), it produces the same warning messages without the rendering problems. Is this problem limited to just this page? You might want to try reloading the page while pressing down the shift key. — Blue-Haired Lawyer t 11:14, 19 November 2010 (UTC)
The problem shows up on several pages. I doubt it's an issue of cache since I arrived at the page by clicking the random article link a couple of times until I got the problem. Taemyr (talk) 12:49, 19 November 2010 (UTC)
Is it only articles with coordinates on them? Does it go away if you disable that F-Secure plugin? Anomie⚔ 16:49, 19 November 2010 (UTC)
Yes, it seems to be only on articles with coordinates. No it does not go away if I disable F-secure. Taemyr (talk) 07:54, 22 November 2010 (UTC)

Link re-writing

I'm working on a test mirror of Wikipedia for my own amusement. Are there handy regular expressions for matching all the links that need re-writing? I've got the /wiki/ ones sorted, but the /w/ ones are proving troublesome. OrangeDog (τ • ε) 21:26, 21 November 2010 (UTC)

For future reference, /w/index.php?title=([^"&]+)&([^&]) -> /wiki/$1?$2 deals with most of them, but the css has to be sent to the index.php directly. OrangeDog (τ • ε) 19:03, 22 November 2010 (UTC)
You need to exclude "?" in the first match or the replacement wont work correctly ;-). In GeoHack we use replace href="/w with what we want to start the url with. — Dispenser 03:22, 23 November 2010 (UTC)

References within navboxes?

Is there a way to add references to navboxes? I get broken output no matter what I try. SharkD  Talk  23:36, 21 November 2010 (UTC)

References and templates in general don't mix well. Also, I think references do not belong in navboxes. — Edokter • Talk • 23:44, 21 November 2010 (UTC)
Why would a navbox need references? OrangeDog (τ • ε) 09:41, 22 November 2010 (UTC)
You didn't give any examples of problems so I don't know what you are trying but I'm not aware of technical problems with references in navboxes. Note however that as for all references, there must be {{reflist}} or <references/> after the reference. Displaying navbox references in articles sounds like a bad idea but the template page for a navbox can use noinclude to display references on the template page below the navbox. PrimeHunter (talk) 10:45, 22 November 2010 (UTC)
Try using the #tag version of ref. -- WOSlinker (talk) 12:21, 22 November 2010 (UTC)
{{#tag:ref|refcontent|name=name}}

Looking at your contributions, the template in question appears to be {{X-COM}} (haven't played those games in a long time). You should have noticed that the error message has a link to the help page at Help:Cite errors/Cite error refs without references, which addresses reference issues in templates (unless your language is set to other than en - English). Many infoboxes and navboxes have references or notes. If the help page is not fully informative, leave some feedback so I can improve it. ---— Gadget850 (Ed) talk 13:32, 22 November 2010 (UTC)

You can also use the "group" parameter and have the ref listed in the "below" section. Femto Bot, (possibly the smallest bot in the world)

Templates for linguistic examples

In an effort to do something against the deplorable state of many entries on linguistic subjects, one thing permanently hampers me, and, I imagine, similar-minded editors: Editing linguistic examples. There is a standardized format to which a linguistic example has to conform:

  • The first line gives the number of the example and the original text in the source language. It should be in italics to offset it optically from the following two lines
  • The second line gives glosses for each word in the first line; each word of the second line is aligned exactly with the corresponding word in the first line.
  • The third line gives a free translation of the example into the meta language.

Here is a rough example:

(4)a.Jedoutequevousayezraison.
1sdoubtthat2phave.subjonctif right
`I doubt that you are right.'

This would be so much easier and faster if there was a template for it. I would make one, but I'm not very experienced with wiki formatting, so I imagine there are people out there who wouldn't take a fraction as long as I would. Anyone? I'd be grateful for any help with this. Watasenia (talk) 09:10, 22 November 2010 (UTC)

Is this a bug with "What links here" (on Commons)?

Hi, try going to Commons:Category:Waterfalls of Chile (a deleted category on Commons) and click the "What links here" link on the sidebar. It (currently) gives 6 hits for pages like Commons:Category:Chile in the 1970s. However I cannot see why it is returning those links. Is this a bug with the way "What links here" works? Jason Quinn (talk) 22:07, 22 November 2010 (UTC)

its some bug with {{Places by decade|Chile|1970}} If you add that to a page it creates a link in the what links here. ΔT The only constant 22:47, 22 November 2010 (UTC)
Confirmed. It has something to do with Commons:Template:Places_by_decade. Still don't know how it's happening. I wrote a note on the talk page there asking pointing them here. Jason Quinn (talk) 23:33, 22 November 2010 (UTC)
commons:Template:Places by decade/main is where the category is coming from; I'm guessin whatlinkshere is just slow at updating in this circumstance, but I'm not fully sure. /ƒETCHCOMMS/ 03:35, 23 November 2010 (UTC)

Formatprice

Per the request "Hi there. Regarding {{Formatprice}}, I would be most grateful if you could change it so that thousands are displayed as e.g. £123,000 rather than "£123 thousand". Thankyou. -Ashley Pomeroy (talk) 22:46, 19 November 2010 (UTC)",[16] would someone please add some of the coding from Template:Inflation to Template:Formatprice to provide that option for Formatprice? Thanks. -- Uzma Gamal (talk) 14:22, 26 November 2010 (UTC)

Extremely slow servers

  Resolved
 – Redrose64 (talk) 19:55, 23 November 2010 (UTC)

Hi, I'm currently logged in through the secure server - something I have never done before (because I believe that it's normally slower than the normal server); however, at the moment it's the only way: the normal Wikipedia server just won't retrieve any page without about 15+ minutes of repeated timeouts and retries - and then the page loads without any skin at all, it looks like it's fully HTML 3.2 compliant, with no 4.0+ features. Windows XP, with both Firefox 3.6.12 and Opera. My internet access to other websites is absolutely fine and normal. So, is there a technical issue? --Redrose64 (talk) 21:05, 22 November 2010 (UTC)

Same for me. I know I'm fairly near you; I suspect an exchange has gone down somehow, as nobody else seems to be having problems. – iridescent 21:11, 22 November 2010 (UTC)
Same here - Chepstow, S Wales. Ghmyrtle (talk) 21:13, 22 November 2010 (UTC)
It's a Europe-wide problem. Danese says: "There is an outage in progress for people accessing Wikipedia over esams. Ops is working to restore normal service" [17] the wub "?!" 21:36, 22 November 2010 (UTC)

Now working OK. --Redrose64 (talk) 19:55, 23 November 2010 (UTC)

Defaultsort and ref tags

If you add a category from within a ref tag, it doesn't seem to obey the defaultsort setting. (see User:WOSlinker/Sandbox2 for an example.) Just wondering if there is a bugzilla item for this. Couldn't find one when I did a search. -- WOSlinker (talk) 22:43, 23 November 2010 (UTC)

I don't know of one, but I did notice that if you put {{DEFAULTSORT}} before the <ref>, then it does seem to obey. Anomie⚔ 01:12, 24 November 2010 (UTC)

Bug in wiki software for autoblocks

Alright, I think I'm going to report this here; I don't know if y'all are knowledgeable about it.

Recently, I blocked CAtruthwatcher (talk · contribs · deleted contribs · logs · filter log · block user · block log); I happened to know his IP is 97.77.103.82 (talk · contribs · deleted contribs · logs · filter log · block user · block log) due to WP:DUCK issues. When I went there, the top read the following:

For 97.77.103.82 (talk ¦ change block ¦ unblock ¦ block log ¦ logs ¦ deleted user contributions ¦ filter log)

Hmm, Mediawiki seems to know the IP is autoblocked, and is telling me this. In fact as an administrator, if anyone goes there right now, they'll see this error.

Now, it's gotten even worse now that I've placed a more recent block on CAtruthwatcher's sock, Achievestudent (talk · contribs · deleted contribs · logs · filter log · block user · block log). Now it's displaying a pink box that anybody, including non-admins can see:

  This account or IP address is currently blocked. The latest block log entry is provided below for reference:

  • (del/undel) 06:05, 6 November 2010 Magog the Ogre (talk ¦ contribs ¦ block) blocked 97.77.103.82 (talk) (anon. only, account creation blocked) with an expiry time of 48 hours ‎ (enforcement of block on User:CAtruthwatcher (confirmed puppet)) (unblock ¦ change block)

    • For those of us that can do math, 6 November + 48 hours != 16 November. The mediawiki is telling the entire world right now that this IP is autoblocked. I've refrained from placing the longer block on that IP to enforce the ban until enough people have been able to see it. Magog the Ogre (talk) 08:46, 16 November 2010 (UTC)

    The pink box sometimes displays older blocks instead of the current one. I once encountered a notice about a 24-hour block from 2005 while checking the contributions for a user who is indefinitely blocked; I can't remember the name of the offending user though. Graham87 10:13, 16 November 2010 (UTC)
    Hmmm, perhaps a similar problem occurs with range blocks as well. Graham87 10:16, 16 November 2010 (UTC)

    What's really got me is the fact that it's occurred twice. The first time I blocked CATruth, within a minute the message went from "block" to "change block" on the IPs page; as such, it was clearly related. And now there's this time. Magog the Ogre (talk) 10:20, 16 November 2010 (UTC)

    Couldn't we just unblock the IP, then immediately re-block it? It would remove the previous blocking box. Eagles 24/7 (C) 21:42, 16 November 2010 (UTC)

    I think you misunderstand. It isn't a problem with blocking. It is simply a problem with the Mediawiki software accidentally giving us more information than it should be, i.e., indirectly confirming that there is an autoblock on this address, i.e., indirectly giving us a way to check if a user is editing from an IP. That's a privacy issue.

    Also, should I take the lack of response here to mean I should post a bug at Bugzilla? Does anyone know of an existing report? Sorry I've never lurked Bugzilla. Magog the Ogre (talk) 00:29, 18 November 2010 (UTC)

    Let's face it. If you don't know what the IP is in the first place you won't be able to link the two. To be able to draw the link, you must (1) know that the IP is likely them, and then (2) check that the IP shows up as blocked after you block the account. But even if MW doesn't show that information, you can still link the two simply by blocking the IP first, then block the acccount, and check to see if an autoblock shows up (if there is an existing block on the IP, there will be no autoblock). So, it doesn't really give you any information you cannot get without relying on the interface change. T. Canens (talk) 09:59, 20 November 2010 (UTC)

    I don't understand; how would blocking the IP allow the admin to see if an autoblock shows up on a user? Only if the user tries to edit and puts an {{unblock-auto}} template up would it become obvious... if the user is smart, then that won't happen. Magog the Ogre (talk) 21:07, 21 November 2010 (UTC)

    If the account has edited recently when you block it, the IP will be autoblocked and you can see that there is an autoblock from Special:BlockList - unless, that is, the IP has been blocked already. So instead of (1) block account and (2) go to the IP you suspect to see if it's been autoblocked, you can simply (1) block the IP you suspect; (2) block the account; (3) see if your block also brings up an autoblock. Both give you the same information, and the latter will always be doable unless we stop showing autoblocks in Special:BlockList (not a good idea). In short, merely knowing that an IP is autoblocked gives you no usable information unless you are already suspecting something to begin with. T. Canens (talk) 01:08, 22 November 2010 (UTC)
    • If the user is currently blocked (indefinitely or for a time-limited period doesn't matter), if you go to the user's very first contributions in their contribs list (prev, not the latest contributions), that account's first block will be displayed in the pink box. If you go to the contribs list of an IP address with block history (even if it currently isn't blocked), if the IP is covered in a range block, the latest entry in that log will show up in the list. IPs with block history that are affected by an autoblock also have their latest block show up in their contribs logs. Also, if you have been blocked yourself, and then unblocked, and your IP is affected by the autoblock (if it's not been lifted), if you are logged in and go to your contribs list, the unblock will show up in the pink block box. HeyMid (contributions) 22:04, 25 November 2010 (UTC)

    Move, - with some links not updates why?

    I decided to try and pitch in by cleanup up some of the backlog in Requested Moves. I ran into a snag on my very first one. The request was to move [[Stars and Stripes Forever]] to [[The Stars and Stripes Forever]]. What could be easier?

    After moving, I decided to check "what links here". Quite a few entries. I wondered if there's a time lag, and sure enough, the next day some were gone. But this page still has many links. Tnxman wonders if there's a bot that runs when things are slow, but I don't think that's the case. I don't even think the change is done by a bot, but I could be wrong. My current thinking is motivated by the fact that in many of these examples, the original wikilink was surrounded by quotes, either the usual double quotes, or pairs of single quotes to make it italic. I'm wondering if there is a rule, not unlike that for correcting misspelling, to avoid updating whenever a term is in quotes? I assume I can use AWB to clean these up, but I'd like to know if my theory is correct, and if there is some other clever way to do the update, or if my original move was done incorrectly.--SPhilbrickT 21:56, 22 November 2010 (UTC)

    Whatlinkshere takes forever to update, often. I wouldn't worry too much. /ƒETCHCOMMS/ 22:05, 22 November 2010 (UTC)
    The links are updated automatically. Moreover, linking to redirects is a normal practice, and the links to them do not need to be updated. Ruslik_Zero 14:03, 25 November 2010 (UTC)!

    Usage of an image redirect

    I recently got File:Old corn mill ar Rhydowen.jpg on the Commons renamed to File:Old corn mill at Rhydowen.jpg. Now I cannot find any way of finding out which articles use the old name! (Note, this is a suggestion for developer action. I have solved this specific question via another route.) — RHaworth (talk · contribs) 15:28, 23 November 2010 (UTC)

    If you go to the redirect page (from the old name) and click "What links here" in the left menu, doesn't that give you what you are looking for? Jason Quinn (talk) 17:12, 23 November 2010 (UTC)
    What links here for File:Old corn mill ar Rhydowen.jpg[18] shows Wikipedia:Village pump (all) (links), Wikipedia:Village pump (technical) (links), User:Anomie/VPA (links). --Uzma Gamal (talk) 14:51, 25 November 2010 (UTC)

    Image only showing as file name

    The following image File:Grand Theft Auto IV Episodes From Liberty City.jpg is only appearing as a link on the file page and isn't appearing in any form on the article pages that it should be on. I've tried purging the page etc but to no avail. Any ideas? - X201 (talk) 12:17, 24 November 2010 (UTC)

    It's working OK for me.--Brianann MacAmhlaidh (talk) 13:01, 24 November 2010 (UTC)
    As a fair use image, perhaps there is coding that prevents the fair use image from being shown within the {{Infobox}}. You may want to post at Template talk:Infobox video game as well. -- Uzma Gamal (talk) 14:47, 25 November 2010 (UTC)

    Right click open Discussion tab into new Internet Explorer tab, How do I

    When I right click a Wikipedia discussion tab from a Wikipedia page, I would like the option to open that Wikipedia discussion page into a new Internet Explorer tab. I'm able to right click linked terms in an article and open them in a new INternet Explorer tab, but can't do that for Wikipedia discussion tabs from a Wikipedia page (and can't do that a Wikipedia page tave from a Wikipedia discussion page). What do I need to change to be able to open a Wikipedia discussion page into a new Internet Explorer tab by right clicking on the tab? -- Uzma Gamal (talk) 14:52, 24 November 2010 (UTC)

    Those tabs are ordirairy links, I can right-click and choose 'open in new tab' without problem. Perhaps there is some other software in the way.I spoke too soon (I'm using Chrome). Internet Explorer is indeed not showing the correct context menu. (Edit:) It seems that the span containing the link text does not trigger the right menu for some reason. However, if you right-click just next or below the text, the menu works correctly. Aditionally, you can also shift-click (open in new window) or ctrl-click (open in new tab). — Edokter • Talk • 15:04, 24 November 2010 (UTC)
    Right clicking seems to work correctly for me (IE8, Windows XP Home edition SP3).Nigel Ish (talk) 17:49, 24 November 2010 (UTC)
    You also could click on the link with your mouse wheel, if your mouse supports it, and this will open the link in a new tab.Smallman12q (talk) 13:06, 25 November 2010 (UTC)
    • Thanks everyone! Right-click on the Internet Explorer tab text, no. Right-click just next or below the the Internet Explorer tab text, yes. That was a surprise. Also, I never knew that I could push on my mouse wheel to cause some action (click on the link with your {mouse wheel). The shift-click (open in new window) and ctrl-click (open in new tab) worked as well. Thank you! This is great. I'm coming back with more questions! -- Uzma Gamal (talk) 14:37, 25 November 2010 (UTC)

    A disappearing post?

    From this diff, a recent posting of mine seems to have disappeared in the next edit. It could have been an accidental deletion by the next editor, but is it also possible that it is a bug? I've reposted it anyway, but I thought I'd ask... AndyTheGrump (talk) 13:51, 25 November 2010 (UTC)

    Seems like the editor had accidentally overwritten it. I doubt it's a bug; if it's an edit conflict, then well, we'll get an edit conflict notice. Rehman 13:57, 25 November 2010 (UTC)
    I'm guessing the other editor got an edit conflict and then saved without resolving it appropriately. I would have asked the editor if I wanted to investigate. PrimeHunter (talk) 14:03, 25 November 2010 (UTC)
    No, it's a bug that happens regularly when two people post to a page close to each other, but just avoiding an edit conflict. Seems to especially happen on large pages. If you search the archives here it appears a number of times. There's a bugzilla for it somewhere, as well. Here's a previous report. Black Kite (t) (c) 00:08, 26 November 2010 (UTC)

    Design assistance

    Could somebody good at these things give a hand with Wikipedia:New contributors' help page/header-redesign (discussion here)? The central boxes new improving, at least to get the alignment right, if not more substantially. Thanks. Rd232 talk 14:21, 25 November 2010 (UTC)

    Need Help With New Article

    I just created the article European Demographic Balance 1939-1947

    How do I update the article to include the blue “Content” box at the top?--Woogie10w (talk) 17:33, 25 November 2010 (UTC)

    The table of contents automatically appears when there are more than three section headings. I'm not sure why you want it now when the only headings are already at the top. PrimeHunter (talk) 18:31, 25 November 2010 (UTC)
    From Help:Magic words you can use __FORCETOC__ to force the display of the table of contents.--Salix (talk): 23:43, 25 November 2010 (UTC)

    MfD nomination of MediaWiki:Talkexists

    MediaWiki:Talkexists, a page you substantially contributed to, has been nominated for deletion. Your opinions on the matter are welcome; please participate in the discussion by adding your comments at Wikipedia:Miscellany for deletion/MediaWiki:Talkexists and please be sure to sign your comments with four tildes (~~~~). You are free to edit the content of MediaWiki:Talkexists during the discussion but should not remove the miscellany for deletion template from the top of the page; such a removal will not end the deletion discussion. Thank you. :| TelCoNaSpVe :| 05:03, 26 November 2010 (UTC)

    Clarify your pronoun antecedents, paleface. ―cobaltcigs 06:21, 26 November 2010 (UTC)

    Inflation

    There are several requests at Template talk:Inflation that are in want of attention. Alexander Gieg, of template Formatprice and Inflation fame, appears to be away for awhile, so I posted here. -- Uzma Gamal (talk) 14:27, 26 November 2010 (UTC)

    Liquid threads treated like a forum/blog?

    Can a liquid threads conversation be embedded in another page? If so how?

    one

    The ability to embed one conversation page in another conversation page.

    two

    In the alternative, can you embed all of the postings with a particular category onto one page?

    For example, the "news" section on http://deadrising.wikia.com/wiki/Dead_Rising_Wiki ?

    There the wikia adds: 

    <bloglist summary="true" summarylength=300 timestamp="true" count=5 type=box>
<title>News</title>
<type>plain</type>
<order>date</order>
<category>News</category>
</bloglist>

    ...to the main page, and all of the postings with the category:news appear on the front page.

    Adamtheclown (talk) 15:21, 26 November 2010 (UTC)

    Hm, this is a very interesting new feature. I've long desired a more suitable method of posting messages on the various talk pages of Wikipedia. However, I'd prefer to see a more fleshed out example first. Would it be possible to place some generic placeholder text on your MediaWiki talk page so we can see what a larger discussion would look like? I agree it would most definitely have to be done on a case-by-case basis though. --Dorsal Axe 16:06, 26 November 2010 (UTC)
    I wish I knew the answers :/. Maybe I can ask the creators to respond here.
    Play with it now on mediawiki.org on your talk page. Simply add {{#useliquidthreads:1}} Adamtheclown (talk) 20:36, 26 November 2010 (UTC)

    Shortened form of Special:Contributions

    Hi everyone,

    Does a shortened form of [Special:Contributions/XX] exist?

    For example, [Wikipedia:User_access_levels] has [WP:UAL]

    Thanks, West.andrew.g (talk) 16:51, 26 November 2010 (UTC)

    For a number of reasons, redirects to the Special: namespace don't work. That means that it isn't possible to create such shortcuts. — Gavia immer (talk) 17:11, 26 November 2010 (UTC)
    I looked at Special:WhatLinksHere/Template:Soft redirect and found sp:new and SP:AF but I'm not sure these or many other of the currently 129 soft redirects [19] in article space are a good idea. If one was created for Special:Contributions then it couldn't include a username. You can create a clickable link with {{Contribs}}, for example {{Contribs|XX}} to give contributions (of course we have a User:XX among 47,320,549 usernames). There is also {{User}} and variations to make additional links. PrimeHunter (talk) 18:16, 26 November 2010 (UTC)

    Marking Edits as Minor Using API

    So I am using the API in-code, and want to mark my vandalism reversions as 'minor'. But the API doesn't seem to agree with this. Here is my method: 

    	public static InputStream edit_revert(long rid, String title, 
			String summary, boolean minor, pair<String,String> edit_token,
			String session_cookie) throws Exception{
		
			// Building post-string is straightforward. Fields known not
			// to contain special characters are not encoded.
		String post_data = "action=edit";
		post_data += "&undo=" + rid;
		post_data += "&title=" + URLEncoder.encode(title, "UTF-8");
		post_data += "&summary=" + URLEncoder.encode(summary, "UTF-8");
		if(minor)
			post_data += "&minor";
		else post_data += "&notminor";
		
		post_data += "&token=" + URLEncoder.encode(edit_token.fst, "UTF-8");
		post_data += "&starttimestamp=" + edit_token.snd;
		post_data += "&format=xml";
		return(api_post.post(post_data, session_cookie).getInputStream());
	}

    Even when I pass "minor=true", it doesn't mark as minor. The undo will commit fully as expected. The "&minor" will be in the HTTP post. It just doesn't mark as minor. Does anyone have an idea why not? Thanks, West.andrew.g (talk) 03:38, 27 November 2010 (UTC)

    For example, this string to the API:
    action=edit&undo=399058219&title=Pilot+%28The+Vampire+Diaries%29&summary=
Reverted+edit%28s%29+by+%5B%5BSpecial%3AContributions%2F67.121.121.63%7C67.121.121.63%5D%5D+
identified+as+test%2Fvandalism+using+%5B%5BWP%3ASTiki%7CSTiki%5D%5D&minor&token=88577c4d41ae
8115fb0ac4deec0b16c0%2B%5C&starttimestamp=20101127034510&format=xml
    Produced this edit. Strange, ehh? West.andrew.g (talk) 03:49, 27 November 2010 (UTC)
    Okay, I've resolved this issue. Turns out you just don't include "&minor" in the HTTP POST, but you need to include "&minor=true". I'd blame this on my own ignorance, but the examples over at the API page indicate this "=true" bit isn't needed. If others find this misleading, maybe we could poke the maintainers over there for clarification on this point? Thanks, West.andrew.g (talk) 04:30, 27 November 2010 (UTC)
    GET parameters (it's GET, not POST) that don't have a value are supposed to be ignored; that's part of the basic spec. I agree that the documentation is needlessly bad, so now that you've reported the correct working value it ought to be changed. — Gavia immer (talk) 04:41, 27 November 2010 (UTC)
    API *write* statements must be POST-ed (edit, rollback, move, etc.). Basic API *reads* use HTTP GET. Thanks, West.andrew.g (talk) 05:24, 27 November 2010 (UTC)
    Well, I've never written anything to query the API, I was just commenting on what you wrote above. Anyhow, as I said I fixed the documentation where I could find it. — Gavia immer (talk) 05:45, 27 November 2010 (UTC)
    I think you have that backwards, Gavia immer. At least when communicating with a PHP script, GET will keep unvalued parameters while a POST will drop them. Anomie⚔ 05:55, 27 November 2010 (UTC)
    We're way off topic at this point - but what I'm saying is that the server is not supposed to interpret such a parameter as meaning anything (by contrast with something like foo= meaning that the parameter foo has the value of the empty string). That's all. — Gavia immer (talk) 06:32, 27 November 2010 (UTC)

    Lines through infoboxes

    I originally asked this at the Help Desk, but it has been suggested I ask here:-

    I see from the archive [20] that this question has been asked before, but without a helpful reply.
    The article I am trying to improve is Henry McCullough where the section line under Biography cuts through the photo in the infobox. I initially thought it related to a recent change in the img_size= parameter, but Franamax has pointed out that the line appears/disappears depending on the pane width.
    I am using IE8 and Vector, the most common browser and most common skin, so this affects many, if not most, readers.
    Can this line be avoided or overcome?
    Arjayay (talk) 10:49, 26 November 2010 (UTC)

    Hi. I think this is mostly to do with the display resolution; what's your resolution? I use Firefox on 1440×900, and it looks fine... Rehman 10:57, 26 November 2010 (UTC)
    I'm on 1152 x 864, I don't know about Franamax. Have played some more, and at full screen there is no line, but like many people I use a "Favorites" sidebar. The line appears and disappears as I vary the pane width.
    Arjayay (talk) 12:17, 26 November 2010 (UTC)
    I followed this from Arjayay's original post on the Help desk. I saw the line too, and have encountered this problem before. When I add the "frame" parameter to the infobox, as I did in this diff, it now displays correctly for me, but I don't know enough about Wikimarkup to know whether this is a dirty fix, and if it causes problems for other browsers and resolutions. Do please revert me if it's a poor edit. Karenjc 12:50, 26 November 2010 (UTC)
    Thanks for trying, but there's no change on my screen - still got the line (yes I've bypassed my cache etc.).
    Arjayay (talk) 13:27, 26 November 2010 (UTC)
    Hmm, ok. I went back in the edit history, and the display problem seems to start with this edit. I tried reverting the changes one at a time, and found that if I removed the "px" after the image size, the line vanished, reappearing if I put it back. I've changed it in the article - how does it look now? (NB: if it works, I have not a clue why!) Karenjc 14:11, 26 November 2010 (UTC)
    Sorry - no change.
    Arjayay (talk) 15:41, 26 November 2010 (UTC)
    This is bugzilla:23663. So turn of IE8 compatibility view mode. —TheDJ (talk • contribs) 18:42, 27 November 2010 (UTC)

    Liquid threads as a user preference?

    If this is better suited at WP:VPP please move.

    When is Wikipedia going to enable the Liquid threads extension? On www.mediawiki.org/ they have this extension and editors can use it if they wish. Here is my talk page:

    http://www.mediawiki.org/wiki/User_talk:Adamtheclown

    I started liquid threads on my talk page simply by adding {{#useliquidthreads:1}}.

    If Wikipedia allowed editors and projects the freedom to use it on a case by case basis, it would not be very disruptive at all.

    what I love about liquid threads is that if someone responds to your comment on another page, you get an alert on the main bar on the top right. Adamtheclown (talk) 15:21, 26 November 2010 (UTC)

    Due to lack of WMF engineers available to work on it, development of LQT has been on the backburner a tad. It's expected that this will be one of the bigger projects earlier next year. A redesign of the interface is planned. —TheDJ (talk • contribs) 13:18, 27 November 2010 (UTC)

    Spurious pending changes message

    For the past 6 hours or so, my watchlist has always been giving me the pending changes message ("There are currently pending revisions to pages on your watchlist"). However, when I click through, there are no pages shown. When I choose to see all the pending changes (not just those on my watchlist), then there are either a few pages not on my watchlist as expected, or no changes at all. This may be another instance of this bug. The comments in the bug imply that it may be a lag/caching problem, but this has been going on for hours. Orange Suede Sofa (talk) 23:18, 26 November 2010 (UTC)

    FWIW, the problem persisted for a few more hours, and then went away. Orange Suede Sofa (talk) 22:50, 27 November 2010 (UTC)

    About PDF files here (on "Download as PDF")

    Is there a way to do either of the following for a downloaded PDF from here? (It's for a user page of mine.)

    1) Get the "align" functions to work: If I have two tables, if I align one to the left and the other to the right, it won't apply to the PDF. Is there a way to do so?

    2) Force a page-break somewhere. If 1 can't be fulfilled, then what I would end up with is that one table is broken up at an undesirable point.

    Aurora Illumina 04:11, 27 November 2010 (UTC)

    This functionality is provided by PediaPress. Best to ask on Help:Books/Feedback, which they actively watch. I note that normal web print already supports this and that most platforms have a method to print to PDF these days, so you might be able to use that in the mean time. —TheDJ (talk • contribs) 12:48, 27 November 2010 (UTC)
    Thanks. Will copy discussion over there. Aurora Illumina 12:54, 27 November 2010 (UTC)

    Broken IUCN links

    In some articles we have links to IUCN Redlist that are broken due to Redlist update in 2008. For example Andean Negrito. I'm not sure about how many broken links there are, but is this something a bot can fix? --MiPe (talk) 10:06, 27 November 2010 (UTC)

    Reference tags

    Hi, can someone look at this recent edit I've made [21] amd work out what I've done wrong, such that the reference tags are not displaying. Thanks. Eldumpo (talk) 17:32, 27 November 2010 (UTC)

    Fixed.[22] Such errors can often by located by searching the source for the last displayed text before things go wrong. In your edit [23] this was "The tournaments featured all 22 teams from Division Three South." PrimeHunter (talk) 18:13, 27 November 2010 (UTC)
    Thankyou. Eldumpo (talk) 20:39, 27 November 2010 (UTC)

    Global suppress display of the fundraising banner?

    Is there any way to suppress display of the scary beggar fundraising banner across all my Wikimedia accounts? Bit fed up of getting a nasty surprise at the top of the page when I click on a link to commons, or wikiquote, or another language wiki and having to go into preferences on each. DuncanHill (talk) 22:50, 27 November 2010 (UTC)

    This blog post by the creators of Adblock describes how to block it (a few different ways). Only works with Firefox, though. Killiondude (talk) 23:03, 27 November 2010 (UTC)

    Safari Extension

    Could some one please create a Safari Extension similar to the Wikipedia Toolbar for Firefox? I find the toolbar very useful, and it would be great if a version for Safari could be created. ~NerdyScienceDude 22:48, 24 November 2010 (UTC)

    If I get time...ǝɥʇM0N0farewell 02:30, 29 November 2010 (UTC)

    Spam Links

    Is there a way we can make auto-confirmed or some other criteria of user, to allow said user to link to black listed links? It is a ridiculous when I can't show a link in a GAC to a website, because other users, use it for spam. CTJF83 chat 00:24, 25 November 2010 (UTC)

    The blacklist only prevents saving working links, and not URLs that don't happen to be linked. You can share relevant links by surrounding them with <nowiki> tags, or just don't put them in brackets and leave off the http:// prefix and the parser won't notice that the URL is supposed to be linked. Of course, that means that someone reading your link will have to copy and paste it if they want to see what it says, but that's only a minor inconvenience. — Gavia immer (talk) 00:32, 25 November 2010 (UTC)
    Wikipedia:Spam blacklist#Requests for delisting links to MediaWiki talk:Spam-whitelist. Administrators can edit the spam whitelist. A permission to directly add blacklisted links to an arbitrary page would be problematic for several reasons even if it was restricted to admins. Don't circumvent the blacklist by adding non-linking url's to article pages. PrimeHunter (talk) 00:36, 25 November 2010 (UTC)
    So basically this can't be done? I suppose using nowiki tags would be fine too. CTJF83 chat 00:44, 25 November 2010 (UTC)
    Outside mainspace you can use nowiki tags if you want to discuss a link. If you want it added to an article then you can make a request at MediaWiki talk:Spam-whitelist. nowiki tags and other circumventions are technically possible in mainspace but the blacklist is there for a reason and shouldn't be circumvented like this. PrimeHunter (talk) 01:52, 25 November 2010 (UTC)

    I think a more reasonable solution would be to have one list for sites which are unsuitable for sourcing encyclopedic content (albethey interesting enough to discuss on talk pages or the village pump, etc.) and one list for urls which should not be present anywhere (because they host malware or otherwise present some danger to viewers). Of course, the software would have to check for this before allowing a cross-namespace page-move. ―cobaltcigs 02:09, 25 November 2010 (UTC)

    Spam occurs in all namespaces. MER-C 02:07, 26 November 2010 (UTC)
    • It's probably impractical (leaving aside whether or not it's desirable). The problem is that the blacklist functions on all subsequent edits to a page - if a privileged user adds it, and a non-privileged user later edits the page to do something else entirely, it will reject their edits until the blacklisted URL is removed. (This allows URLs to be blacklisted when already in articles, and forces them to be "weeded out" over time). As a result, this would have to be substantially more complicated than a simple per-user override - you'd need to have each URL recorded as "approved" or "non-approved" from the point of addition onwards, and I can't see an easy way to do that. Shimgray | talk | 15:29, 28 November 2010 (UTC)

    Template help (dabbing)

    The template Template:History of Georgia generates a link to History of Georgia, which is a dab page. It should link to History of Georgia (country). Help! Thanks, DuncanHill (talk) 11:20, 28 November 2010 (UTC)

    {{Country history}} says to use the linkoverride= field. DMacks (talk) 11:25, 28 November 2010 (UTC)
    Thank you, that fixed it. DuncanHill (talk) 11:53, 28 November 2010 (UTC)

    Multi-language image descriptions

    I noticed this on the picture of Earthrise, but I'm aware this will relate to every Commons-hosted file with multiple language descriptions (and/or POTD tagging).

    If you view the local file description page File:NASA-Apollo8-Dec24-Earthrise.jpg, as a logged-in or logged-out user - you see all 16 language versions of the POTD caption (and the English one twice). You also see 5 different language file descriptions. However, as a logged-in Commons user accessing commons:File:NASA-Apollo8-Dec24-Earthrise.jpg I only see the English language file description and POTD caption.

    It would be extremely useful if we could make the local file page display only the English language info in cases like this. en.wiki users only care (normally) about the English description, so getting rid of all the foreign language "junk" would make the page much more valuable for Wikipedia readers.

    Is there any way this could be done?--Nilfanion (talk) 12:05, 28 November 2010 (UTC)

    Can you get image File:Vc116 to appear between the table and the Listen box

    On Flywheel, Shyster, and Flywheel, can you get image File:Vc116 to appear between the table and the Listen box so they all lie in the same row. Thanks. -- Uzma Gamal (talk) 15:20, 28 November 2010 (UTC)

    You could do it using a table, but I don't think such layout is a good idea, we don't have to fill every bit of empty space. Svick (talk) 15:33, 28 November 2010 (UTC)

    Bizarre sig error

    In this diff I removed a stray reftag which was blanking the rest of the page after it. As you will see, as a result my signature got inserted where Bugs' sig should have been (I fixed this with my next edit). Is this a known bug? DuncanHill (talk) 15:39, 28 November 2010 (UTC)

    Yes, it is a known bug that signatures and other similar stuff don't work inside <ref> tags, which was what caused the problem you are describing. Svick (talk) 15:56, 28 November 2010 (UTC)
    OK, thanks. DuncanHill (talk) 01:17, 30 November 2010 (UTC)

    iPhone App Problem

    When I go a search for Weird Al and select his page, the page does not load and then crashes a few seconds after. Able to repeate. Sophie (Talk) 13:23, 29 November 2010 (UTC)

    iPod touch version of Wikipedia

    Wikipedia is not working on iPod touches. (the normal version of Wikipedia. Saying something about a Cascading Style Sheets error. Thanks! Perseus, Son of Zeus 13:56, 24 November 2010 (UTC)

    Can you provide the exact error it is giving? — Edokter • Talk • 14:00, 24 November 2010 (UTC)
    On my iPod Touch I find that both the regular and the mobile versions of Wikipedia are working, as of this moment. EdJohnston (talk) 15:49, 25 November 2010 (UTC)
    same with me. Maybe should should also ask apple Breawycker (talk) 00:36, 30 November 2010 (UTC)
    • Working fine for me too, as Breawycker said try contacting Apple, they should be able to help you. Or if you have an Apple ID you can also try asking people on the Apple Forums where I hang out. - Dwayne was here! ♫ 01:41, 1 December 2010 (UTC)

    glossika.com

    http://safebrowsing.clients.google.com/safebrowsing/diagnostic?site=http://www.glossika.com curently say that glossika.com is "suspicious". Could some computer security expert check this? If it is confirmed, then I will ask for a bot. Visite fortuitement prolongée (talk) 22:25, 29 November 2010 (UTC)

    • Website is inactive. The message that comes up is that the hosting service has suspended the site. Nothing to look at. /ƒETCHCOMMS/ 16:07, 30 November 2010 (UTC)
    In my opinion, any link which gets added to a page should be checked against the Google Safe Browsing lists. Consider the malware attack which occured on the German Wikipedia a while back. I've run a comparison as part of my research, and I'd estimate ~5 links a day get added that are on the list. While FireFox integrates the SB lists nicely, those on other browsers might not be so fortunate. Perhaps a bot with this kind of functionality should be constructed (or I could trivially build it into the edit-processing of my WP:STiki tool). Thoughts? Thanks, West.andrew.g (talk) 18:07, 30 November 2010 (UTC)
    We could just add more listed attack sites to the external link blacklist. We already block major URL shortening services due to the potential for a malware attack. /ƒETCHCOMMS/ 02:00, 1 December 2010 (UTC)
    In the case of the Google lists, that isn't possible. Google only provides a MD5 hash of the blacklisted URLs. Every URL encountered needs to be hashed and compared against the list (and at multiple granularity -- i.e., just the domain, 1-level deep into the directory path, n-level deep in the path, etc.). A bit too much computation for the blacklist extension, IMHO. Thanks, West.andrew.g (talk) 17:46, 1 December 2010 (UTC)

    User Scripts

    Hi :). About a month ago, all my user scripts in my monobook.js stopped working (bar Twinkle, Friendly and HotCat - all enabled through my preferences). I checked my JavaScript, it was on, Firefox is up to date (3.6.12), and after I tried to install every script I always purged and bypassed my cache. Especially now with the December drive coming up, I'd find proseize useful, and I'm really starting to miss FurMe and AFC helper, so if anyone knows how to get them working I'd be very grateful. Thanks :) Acather96 (talk) 16:51, 30 November 2010 (UTC)

    Are you running the vector skin? they will need to be moved to vector.js in that case. Peachey88 (T · C) 21:37, 30 November 2010 (UTC)
    I'm not sure why exactly it stopped working, but if you follow these two edits I made to a copy of your monobook.js, it should start working again. Svick (talk) 22:41, 30 November 2010 (UTC)
    Wow Svick, thanks so much :) Acather96 (talk) 07:36, 1 December 2010 (UTC)

    Email from a network engineer offering to donate time

    Ticket:2010111310002415 is from a network engineer offering to donate his time and coding abilities to help Wikipedia. Where should I refer him?--Chaser (talk) 23:13, 30 November 2010 (UTC)

    Danese Cooper or irc://freenode/mediawiki ΔT The only constant 23:21, 30 November 2010 (UTC)
    Or How to become a MediaWiki hacker. Graham87 06:55, 1 December 2010 (UTC)
    Thanks all.--Chaser (talk) 15:19, 1 December 2010 (UTC)

    How do I hide this bloody "A personal appeal from Wikipedia founder Jimmy Wales" banner?

    If I turn Javascript off, it doesn't appear, but otherwise it comes up after the page loads. --NE2 23:51, 10 November 2010 (UTC)

    There's a barely-visible X in the upper-right corner. Or you could try user CSS something like #centralNotice .siteNoticePic { display:none; } Anomie⚔ 01:02, 11 November 2010 (UTC)
    A more general solution using user CSS is:#siteNotice {display: none !important;} which will hide the entire sitenotice permanently. I've never suffered any injury from doing so, and the ratio of useful information to nonsense has always been pretty bad with that feature. — Gavia immer (talk) 01:21, 11 November 2010 (UTC)
    Someone higher-up fixed something, since it's gone again. I have something in my user CSS/JS that had successfully hidden it until today. --NE2 02:56, 11 November 2010 (UTC)

    The script loads from Special:BannerController, so you could have Adblock Plus target urls containing that title. ―cobaltcigs 09:32, 14 November 2010 (UTC)

    By the way, the X isn't much use, because it'll just come back up the next time you come on. Enigmamsg 00:57, 18 November 2010 (UTC)
    by far this is the most annoying banner I have ever had to deal with or seen. Adamtheclown (talk) 18:23, 19 November 2010 (UTC)

    Any way of blocking it on the page itself? It clashes with a bunch of Top icons that I have on my user page, and any page with icons or geo-coordinates (see Fucking, Austria for example). This wasn't exactly planned so well... — Deontalk 07:14, 20 November 2010 (UTC)

    Special:Preferences -> Gadgets -> "Suppress display of the fundraiser banner". T. Canens (talk) 10:02, 20 November 2010 (UTC)
    You, T. Canens, are my hero. --Bobak (talk) 19:24, 2 December 2010 (UTC)

    Since we all benefit from the money that comes in from this banner, perhaps we could put up with it for a few weeks every year. --Bduke (Discussion) 11:56, 20 November 2010 (UTC)

    And if I pay I still have to look at this obnoxious banner? I paid/donated in the past but won't do this year. Largely due to this - imho - too intrusive ad. And I also don't want to see jimmys face every time I look something up in wikipedia. —Preceding unsigned comment added by 91.67.232.210 (talk) 10:59, 24 November 2010 (UTC)

    IMHO, the banner is not doing good; it's really actually pushing away readers (and reducing any possible $$$). I actually force myself to sign in when I lookup random articles, just to get that horrible (no offence to Jimmy or Wikipedia) banner off. I really really think something has to be done to make it more user friendly. Rehman 13:51, 24 November 2010 (UTC)
    To 91.67.232.210: all you need do is create a login, then when you are logged in you get an option to hide that banner. --Redrose64 (talk) 19:29, 24 November 2010 (UTC)
    I like the new banners with real editors shown, more attractive and more interesting.
    We see Jimbo all the time (often in rather unflattering pictures which users make), it seems rather egotistical of him that he be plastered atop 3 million plus pages. Jim had to relinquish his user rights this year, making him less of a leader, and more of a figurehead. The older and more mature wikipedia gets the less wikipedia is about Jimbo, the more wikipedia is about the individual editors as a whole. So these new banners showing one editor are more the future face of wikipedia than pictures of Jimbo. Adamtheclown (talk) 00:19, 25 November 2010 (UTC)
    Ugh. Is it really necessary to make the banner cover the entire window and force me to scroll down? ----Divebomb is not British 17:35, 25 November 2010 (UTC)
    Click on the "X" button top right --Redrose64 (talk) 21:00, 25 November 2010 (UTC)
    I did. I was just asking whether it is really necessary to make it so huge. ----Divebomb is not British 07:20, 26 November 2010 (UTC)

    Where does it say what they need the money for?Chrisrus (talk) 07:40, 26 November 2010 (UTC)

    Basically? Staff and servers. Slightly longer version? see the FAQ. Even longer version? See the 2010-11 Annual Plan. Ale_Jrbtalk 11:16, 27 November 2010 (UTC)

    Seriously. For how long will this campaign last? The next time, a less intrusive and more productive ad would be appreciated. Or is it a permanent feature? Hexmaster (talk) 22:30, 1 December 2010 (UTC)

    If the intrusiveness of this were not sufficient to deter me from any thought of donating while this person was in charge, the ostentatious, brazen revisionism of the whole "founder" lark certainly would. Perhaps we can have a sponsored WikiCoup as the next fundraising event? Smartiger (talk) 05:32, 2 December 2010 (UTC)

    Request for search data

    I have another one. Ticket:2010110810009795 is a request for search data, specifically how much time users take between searches. Any ideas?--Chaser (talk) 15:19, 1 December 2010 (UTC)

    Try the wiki-research mailinglist. —TheDJ (talk • contribs) 20:19, 2 December 2010 (UTC)
    Oh, I had no idea about the existence of that mailing list. Thanks! When people ask for (private) data for research, I usually forward the email to a tech staffer. Killiondude (talk) 21:35, 2 December 2010 (UTC)

    Maybe a bug?

    I'm not sure if this a bug, or an error with my Wikipedia account or what. First off, I feel it might be worth mentioning that I Usurped this nickname a couple of months ago.

    I browse the recent changes list a lot and hand out warnings to vandals of our fine encyclopedia, a lot of them being IP addresses. It seems, though, that everytime I edit and add a warning to an IP address' talk page, Mediawiki adds that talk page to my watchlist. I've checked my settings for my watchlist, which I can verify here: http://img602.imageshack.us/img602/8839/wikipediasettings.png are set so that only pages I create are added to my watchlist, but I typically don't create these talk pages. For example, when I added this edit to User_talk:195.195.4.202 page, it added that page to my recent changes. What might be going on? JguyTalkDone 02:58, 2 December 2010 (UTC)

    Are these manually-added warnings, or warnings added by a script such as Twinkle? Some scripts add the talk pages of users whom you warn to your watchlist. I'm not sure why this happens, but you could ask on the talk page of the script you are using. If these are manually added warnings, is the "Watch this page" box checked? Intelligentsium 04:08, 2 December 2010 (UTC)
    hi, thanks for the speedy reply. Some warnings have been from my Twinkle install on this laptop, some warnings have been from my desktop, which does not have Twinkle (IE 8). I'll keep an eye out on if this makes a difference. thanks again. JguyTalkDone 16:00, 2 December 2010 (UTC)

    New username patrol???

    I was aware of WP:NPP, but this is... rather odd: [24] (and very old...) 2 admins blocked this user within the same minute the account was created. Does anyone know where said admins monitor new accounts? (and if you're wondering about where I got that username from, see the image for WP:AUTOBLOCK). --NYKevin @314, i.e. 06:32, 2 December 2010 (UTC)

    There is the User Creation log. Looking at the block log for around the time of that block, there appeared to be a user or number of users creating accounts with names asking to be blocked. --Kateshortforbob talk 11:57, 2 December 2010 (UTC)
    Special:Log/newusers is the way to navigate there easily. There are bots (on-wiki & on IRC) that report those as well as people who peer over the creation logs. Killiondude (talk) 21:37, 2 December 2010 (UTC)

    Font?

    Is it just me, or did the default Wiki font change recently. I'm using a different computer than usual. IE8 browser. I typically use firefox. Any help is appreciated. Thanks, Hamtechperson 13:39, 2 December 2010 (UTC)

    The default font is set to 'sans-serif', which ususally in Arial being used. But another computer or browser may use a different font. — Edokter • Talk • 13:53, 2 December 2010 (UTC)

    Pyrospirit metadata script

    Pyrospirit metadata script is a User interface gadget provided to each Wikipedia user (all 13,515,834 of them) through their My preferences. It is gaget option "Display an assessment of an article's quality as part of the page header for each article. (documentation)". Pyrospirit last posted 18 May 2010 and posts at User_talk:Pyrospirit/metadata after that date have gone unaddressed. Would someone please continue Pyrospirit's good work with the Pyrospirit metadata script and address the posts at User_talk:Pyrospirit/metadata. Thanks. -- Uzma Gamal (talk) 14:07, 2 December 2010 (UTC)

    I have put the page on my watchlist, but I have no idea if and when I'll get around to it. Someone else is free to be a lot quicker than I am :D —TheDJ (talk • contribs) 20:27, 2 December 2010 (UTC)

    Wrong placement in Speedy deletion category

    Several unrelated pages are present in Category:Speedy deletion, mostly Turkish footballers. They probably got there because of a wrong edit to some template, but i can't find which one. Can anyone help me find it?

    More generally, is there a convenient way to find such things in the future without having dive into each transcluded template?

    Thank you. --Amir E. Aharoni (talk) 14:40, 2 December 2010 (UTC)

    I just used Special:RecentChangesLinked on one of the affected pages to identify the offending template as probably being {{Fenerbahçe SK 100th Year Champion Squad}}. Algebraist 14:59, 2 December 2010 (UTC)
    Sometimes there are hints. In this case the badly edited template {{Fenerbahçe SK 100th Year Champion Squad}} was itself in Category:Speedy deletion (deleted admin only diff: [25]). I have fixed it. PrimeHunter (talk) 15:00, 2 December 2010 (UTC)

    Wikipedia:Syndication bugs

    Several editors, including me, have noted various problems at Wikipedia talk:Syndication, but it does not appear that this page is being watched by many people that can fix the problems, or even answer the questions. I'd like to bring it to your attention (and in particular, my recent reported problem there). --Piotr Konieczny aka Prokonsul Piotrus| talk 16:27, 2 December 2010 (UTC)

    Image thumbnail problem

    Right thought I could solve this myself and have failed in epic proportions. File:DohaKhalifaStadium.jpg was uploaded in Feb 2010. Then some new images (un-free with no change of license) were uploaded over these. I reverted to the old version and deleted the non-free intermediate revisions. However, it keeps displaying the thumnail of the deleted image and even did so after I completely re-uploaded the correct image. No deletion/restoration combo seems to have worked (and any admin will see I've tried a few!). I can never get the top (most recent) thumnail to display the correct image. Any ideas? Rambo's Revenge (talk) 16:42, 2 December 2010 (UTC)

    Cache? it looks OK to me. --ClubOranjeT 16:50, 2 December 2010 (UTC)
    Not cache. Could you check the Khalifa International Stadium infobox then. That still displays the deleted thumbnail despite purges/nulls. Which version does it display for you, the arial view or the one with from the stand with fireworks? Rambo's Revenge (talk) 16:54, 2 December 2010 (UTC)
    Infobox has the one with from the stand with fireworks. --ClubOranjeT 17:19, 2 December 2010 (UTC)

    Banners refuse to close

    Chrome 6.0.472.63, Win XP. I've always been able to hide the banners by hitting the [x] but it's not working this time. This is with the "Lilaroja" banner. Any ideas? /ƒETCHCOMMS/ 18:50, 2 December 2010 (UTC)

    Using Firefox. Hopefully it's a quick fix, because it's driven me mad. --Jezebel'sPonyobons mots 18:59, 2 December 2010 (UTC)
    • Me too. And other banners are popping up as well, but those are able to be closed. It used to be they only popped up when you logged on, which was OK... - The Bushranger Return fireFlank speed 19:00, 2 December 2010 (UTC)
      • Aha. In the top subject of the page, T.Canens says:
    Special:Preferences -> Gadgets -> "Suppress display of the fundraiser banner"
    (edit conflict) I'm encountering the exact same problem with one banner from "Lilaroja". No matter how many times I click "close" it doesn't go away. I'm using the more recent version of Firefox on Windows 7. —Farix (t | c) 19:02, 2 December 2010 (UTC)
    • The banners are back to normal now for me. And yes, I know there's a gadget, but I like seeing what banners they're using :) TheFarix, maybe try again? If you keep reloading, you're bound to get one of the banners that should close. /ƒETCHCOMMS/ 19:02, 2 December 2010 (UTC)
    No matter how many times I try, they still one stay closed, with the "Lilaroja" banner not closing at all. —Farix (t | c) 19:04, 2 December 2010 (UTC)
    I'm having the exact same problem as the folks above. Firefox 3.6.12. - Burpelson AFB ✈ 19:08, 2 December 2010 (UTC)

    Me too. This is killing me. I even checked the preference gadget to hide them completely and the obnoxious things are still there. Gigs (talk) 19:10, 2 December 2010 (UTC)

    Back to borked for me as well. The issue I think is that one single Lilaroja banner has some coding issue which does not let it be hidden. I hid the other banners, and that one still pops up half the time. Perhaps wikimedia-tech? /ƒETCHCOMMS/ 19:11, 2 December 2010 (UTC)
    It's definitely the Lilaroja banner that's causing the problems. I can close any of the others, but then she pops up. Nothing against poor Lila personally, but it's driving me to distraction. --Jezebel'sPonyobons mots 19:13, 2 December 2010 (UTC)
    You can permanently disable them by adding to Special:Mypage/skin.css -
    #centralNotice  .siteNoticePic,
{
    display: none;
}
    Of course, doesn't help people who like to see what they're using. –xenotalk 19:15, 2 December 2010 (UTC)
    Tried to wp:clear your cache? LeadSongDog come howl! 19:28, 2 December 2010 (UTC)
    I'm having this problem as well, and to be honest, Xeno's solution is unacceptable. This needs to be fixed, or the banner needs to be taken down until it can be fixed. --Shirik (Questions or Comments?) 19:31, 2 December 2010 (UTC)
    More of a workaround than a solution. I've made a posted at meta:Talk:Fundraising 2010/Messages#Issue with Lilaroja banner refusing to close. –xenotalk 19:36, 2 December 2010 (UTC)
    Oh I know, and what I said wasn't directed at you but more of a "I hope someone's still looking at this and hasn't given up because of this workaround being posted" --Shirik (Questions or Comments?) 20:06, 2 December 2010 (UTC)
    I even blocked all images from bits.wikimedia.org trying to get rid of this banner; only setting the gadget worked. I hope someone at the Foundation is paying attention because problems like this will only discourage editors to contribute their names & faces in future fundraisers. (And I too hope no one will take out their frustration over this screwup on Lilaroja.) -- llywrch (talk) 19:43, 2 December 2010 (UTC)
    • The fundraising staff team is aware and is currently addressing this issue, which should be resolved shortly. PeterSymonds (talk) 19:47, 2 December 2010 (UTC)
    Fixed, shouldn't happen again in the future, and thanks for those that pointed this out to us. If you use IRC, you can almost always find at least one staff member in #wikimedia-fundraising. Keegan, Wikimedia Fundraiser 2010 (talk) 20:43, 2 December 2010 (UTC)
    Thank god that's gone. That was horrible. AtticusX (talk) 20:36, 2 December 2010 (UTC)
    Thanks Keegan and everyone else; I didn't have time to hop on IRC or I would have poked someone faster. /ƒETCHCOMMS/ 01:15, 3 December 2010 (UTC)

    ProveIt reference management tool released

    Georgia Tech has just completed a new tool for managing Cite.php based references, ProveIt. It provides a convenient GUI that eliminates the need for beginners to learn Wikipedia syntax before assisting with referencing. It supports adding and editing references, as well as adding additional citations to existing references. Additional information is available on ProveIt's website.

    We encourage people to use it and let us know your feedback. Bug reports can be submitted here, or you can comment on our talk page. Superm401 - Talk 04:03, 3 December 2010 (UTC)

    Conscientiously reading "recent changes" syndication

    I'm goofing with a little personal program (using Python and Feedparser) to extract some statistics about edits. I'll be retrieving the XML (or ATOM) feed of Special:RecentChanges. Looking at what I'm getting from that gives me a few questions about the correct, conscientious ways of doing this: 

      while True:
    get_changes_since(last_time -1minute)
    last_time = now
    process_received_stuff()
    wait(5 minutes)
    (where I have to do that -1minute because the previous fetch won't have retrieved all of the edits for the last minute)
    Is there a smarter way to do this?
    • The above scheme means that subsequent fetches will see some of the edits detailed in the previous fetch, so it'll be my responsibility to skip duplicates. It looks like I can use the {{{1}}} field as a unique identifier (from e.g. the entry id url "http://en.wikipedia.org/w/index.php?title=Ibadan&diff=400219658&oldid=prev"). Is this a safe assumption?
    • In practice how frequently should I retrieve the feed? I don't care at all about real-time performance (no-one will be actively watching the program, and it won't respond in any way), so I don't care if it updates every 10 minutes or every hour. But as there isn't a "to" field, I end up always retrieving the last N edits, so I might as well get it once every few minutes. Ideally I'd use a query string that retrieved a recent-changes served from cache rather than making the DB do a unique query just for me, but I can't think of a way of writing that query string. It seems to me that, unlike article views, the RecentChanges page (and the corresponding feeds) won't be cached.
    • The above scheme is a bit ad-hoc, in that it doesn't react to the wiki getting busier or quieter. And doing all this is a bit of work. Has someone already coded a library (ideally in Python) that simplifies the fetch in a smart way: that is, that it conscientiously queries the feed at an adaptive rate and and yields a complete, unique stream of entries?
    • Or am I overthinking, and overlooking a really obvious way of getting a unique update stream without all this extra processing?

    Thanks. -- Finlay McWalter ☻ Talk 23:49, 2 December 2010 (UTC)

    There's an IRC feed of recent changes (or at least there used to be) at #en.wikipedia. Maybe that would be useful? the wub "?!" 12:07, 3 December 2010 (UTC)
    Thanks, yes, that's easier and lighter-weight. Is the format officially documented somewhere - I've mostly figured it out, and it's vaguely like the HTML rendering of the same data described at m:Help:Recent changes, but I'd be nice if there was a real specification, so I could be confident I'd handle rare occurrences that I haven't seen. -- Finlay McWalter ☻ Talk 19:31, 3 December 2010 (UTC)

    See Mediawiki API: Recent changes and get back to the API team if there is something you want to do that isn't covered there. Note particularly the rcstart and rcend parameters. --TS 00:41, 4 December 2010 (UTC)

    Infobox writer influences parameter, show/hide

    In the Infobox writer template, there is a parameter called influences. When you use the parameter, the template apparently creates a show/hide toggle icon. See Justin Cronin. The template documentation doesn't say anything about it. Any way to not show the toggle, meaning always show the influences? I can see wanting to hide influences when the list is long, but, otherwise, it seems silly to hide them.--Bbb23 (talk) 01:56, 4 December 2010 (UTC)

    Where to report this bug?

    I tried uploading a file with a colon (:) in the name on commons; it didn't disallow it, but I got this really unfriendly message:

    Internal error

    key '2muj1nywb867g7yba09qnnr99h4qos5.' is not in a proper format

    Backtrace:
    #0 /usr/local/apache/common-local/wmf-deployment/includes/upload/UploadBase.php(557): UploadStash->stashFile('/tmp/phpnXpjgd', Array, NULL)
    #1 /usr/local/apache/common-local/wmf-deployment/includes/upload/UploadBase.php(569): UploadBase->stashSessionFile(NULL)
    #2 /usr/local/apache/common-local/wmf-deployment/includes/specials/SpecialUpload.php(322): UploadBase->stashSession()
    #3 /usr/local/apache/common-local/wmf-deployment/includes/specials/SpecialUpload.php(413): SpecialUpload->showUploadWarning(Array)
    #4 /usr/local/apache/common-local/wmf-deployment/includes/specials/SpecialUpload.php(167): SpecialUpload->processUpload()
    #5 /usr/local/apache/common-local/wmf-deployment/includes/SpecialPage.php(560): SpecialUpload->execute(NULL)
    #6 /usr/local/apache/common-local/wmf-deployment/includes/Wiki.php(254): SpecialPage::executePath(Object(Title))
    #7 /usr/local/apache/common-local/wmf-deployment/includes/Wiki.php(64): MediaWiki->handleSpecialCases(Object(Title), Object(OutputPage), Object(WebRequest))
    #8 /usr/local/apache/common-local/wmf-deployment/index.php(117): MediaWiki->performRequestForTitle(Object(Title), NULL,
    Object(OutputPage), Object(User), Object(WebRequest))
    #9 /usr/local/apache/common-local/live-1.5/index.php(3): require('/usr/local/apac...')
    #10 {main}

    Is this a known bug? If not, how do I report it? Magog the Ogre (talk) 00:34, 2 December 2010 (UTC)

    bugreport submitted. Thanx for the proper description. Can you perhaps add which browser/version you were using ? might be handy in case it is a Javascript bug. —TheDJ (talk • contribs) 20:22, 2 December 2010 (UTC)

    Google Chrome for Ubuntu, 32 bit. Magog the Ogre (talk) 08:56, 4 December 2010 (UTC)

    Help:Searching

    How come Help:Searching isn't linked from Special:Search? I thought it used to be. Rd232 talk 17:47, 3 December 2010 (UTC)

    It's linked from MediaWiki:Searchmenu-exists and MediaWiki:Searchmenu-new. The only way not to get the link is to go to Special:Search with no searchterm. Algebraist 17:53, 3 December 2010 (UTC)
    I see. That doesn't seem ideal - those users should see it too. It could be a link like the one at Special:Contributions - small, top right, consistently there in addition to other references to Help. Rd232 talk 02:04, 4 December 2010 (UTC)
    As far as I know, there is no persistent MediaWiki message that we can abuse for this purpose. That's why MediaWiki:Searchmenu-exists and MediaWiki:Searchmenu-new are used right now. —TheDJ (talk • contribs) 21:57, 4 December 2010 (UTC)
    I think there should be a persistent MediaWiki message to enable this. Do we agree it's worth filing a bug? Rd232 talk 01:26, 5 December 2010 (UTC)

    Feedback

    I noticed while browsing GFAJ-1 that there's a feedback section at the bottom of the page. I've never encountered this before and I feel like I've missed something here. Could someone explain what this is all about, and why only very few articles (I assume) seem to feature this function? --Dorsal Axe 18:53, 3 December 2010 (UTC)

    I think it's part of this. – ukexpat (talk) 19:21, 3 December 2010 (UTC)
    That must be it. Thanks. --Dorsal Axe 12:18, 4 December 2010 (UTC)

    Mobile Site Formatting Problem on Palm Pre

    I have a Palm Pre, running Palm's WebOS. Since Thursday (12/2) morning, wikipedia mobile pages have been coming up with different formatting. Instead of the usual, easily readable, single columb. The pages have been spread out quite a bit wider, requiring zooming and subsequent side-to-side scrolling. In order to stop the inevitable, "it must be your fault" responses, I've double checked, and it is the mobile site at which I'm looking. Furthermore, a quick web search reveals that I'm not the only Palm user that has noticed this, so it's not relugated to an error on my phone. Palm's browser doesn't have formatting options, so it's not that, besides, I didn't change anything from when it was displaying properly to when it stopped. I'm writing this in the hopes that the powers that be that control the actual Wikipedia mobile site will find out, and fix this. —Preceding unsigned comment added by 70.226.217.101 (talk) 03:07, 4 December 2010 (UTC)

    Filed as bugzilla:26239 —TheDJ (talk • contribs) 21:46, 4 December 2010 (UTC)

    Ugly link in printable version

    The wikitext over {{ma|3500}} renders as "over 3,500 million years ago" in an article, and as "over 3,500 (http://toolserver.org/~verisimilus/Timeline/Timeline.php?Ma=3,500) million years ago" when using the sidebar "Printable version". Unfortunately, the template is used frequently in some articles (example Evolutionary history of life or printable) and the result is unusable. The issue was raised here. Is there some way of fixing that in {{ma}}? Johnuniq (talk) 03:17, 4 December 2010 (UTC)

    Replaced deleted class plainlinksneverexpand with nourlexpansion to prevent the URL showing when printing. ---— Gadget850 (Ed) talk 04:37, 4 December 2010 (UTC)
    And I see you've done that. Excellent, thanks. Johnuniq (talk) 07:22, 4 December 2010 (UTC)

    Bot stalking!

    Hello, I have a little problem. Everytime I post, and sign (either with the button or manually) SineBot seems to unrecognize it and automatically sign for me later. What's happening? I'm quite an experienced user at es:wiki and this never happened to me before. Fernando 15:04, 4 December 2010 (UTC) —Preceding unsigned comment added by Fernando (talk • contribs)

    Looks like your sig is linking to User:Fernandopascullo but your actual user name is User:Fernando. That's why. ♫ Melodia Chaconne ♫ (talk) 15:09, 4 December 2010 (UTC)
    (edit conflict) Yes, Melodia is right. You can change your signature at Special:Preferences. Rehman 15:12, 4 December 2010 (UTC)
    *facepalm* Thanks guys! Fernando 15:39, 4 December 2010 (UTC)

    Invalid HTML in fmbox

    The script User:Ucucha/autofc.js, which I just created, apparently is unable to edit WP:Featured article candidates because of some invalid HTML in the edit notice for that page. The error seems to stem from Template:Fmbox. Does anyone know how to fix this? Ucucha 17:01, 4 December 2010 (UTC)

    It turns out the problem at Fmbox is unrelated (I believe it relates to empty ul's created by Vector). My problem was instead caused by something somewhere that caused the HTML for the FAC editnotice to first close the < td>, and only then the < li> and < ul> for "Finally, to help resolve...". I solved this by adding an nbsp, which solves the problem with my script. However, it would be better if the underlying problem could be resolved. I reproduced it here at User:Ucucha/sandbox. Ucucha 17:26, 4 December 2010 (UTC)

    Search Everything

      Resolved

    Why aren't media files included when you search with "everything"?Smallman12q (talk) 17:47, 4 December 2010 (UTC)

    They are included. What makes you think otherwise? PrimeHunter (talk) 17:58, 4 December 2010 (UTC)
    Just checked again and you're right...must've made a mistake.Smallman12q (talk) 20:19, 4 December 2010 (UTC)

    Is there documentation for the recent changes IRC feed?

    Sure I can look at it, and in the limited exposure I get pick out some format that will allow me to read some of it. But is there a full documentation, describing the format of anything that the rc-pmtpa will throw at me? 930913 (Congratulate/Complaints) 20:33, 4 December 2010 (UTC)

    Table Help

    Please look at the Anne Hathaway article, in particular the 2009 row for the Simpsons. In the Role column, I want to vertically top align that cell so that Jenny appears at the top. I looked at various stuff intended to help me understand tables to no avail. How do I do it?--Bbb23 (talk) 23:34, 4 December 2010 (UTC)

    You can use the HTML style attribute (preferred over the deprecated valign attribute) to add code like this: style="vertical-align:top" to the row, like this:
    |-style="vertical-align:top"
|Content
|Content
|-
    Why doesn't the default style for wikitables include this though? It seems to me that would be a more useful default. --Mepolypse (talk) 00:11, 5 December 2010 (UTC)
    Could you please be more specific as to how to add it to the Hathaway entry? Sorry.--Bbb23 (talk) 00:14, 5 December 2010 (UTC)
    (edit conflict) Sorry, I made a typo and also improved the suggestion above. I'll apply this change to the first row in the article as well, so you can see it in use. --Mepolypse (talk) 00:17, 5 December 2010 (UTC)
    Got it, thanks, I've removed your change from the article and incorporated it (along with some white space) to the relevant row. Looks much better. Hopefully, the editor who found it confusing before these changes will leave it alone. Thanks again.--Bbb23 (talk) 00:27, 5 December 2010 (UTC)
    Oh, I see what you were trying to do now. I think what you want is two rows with a rowspan attribute on joined cells. I've made that change, feel free to revert if you don't like it. --Mepolypse (talk) 00:50, 5 December 2010 (UTC)

    Pulling data from external API

    What I would like to do is scrape a single integer from an external site (via an API), and make it visible within a Wikipedia article. In particular, I would like to update WP:STiki with the number of vandalisms it has undone to date (which requires a DB query on my local machine). Is this possible?

    Obviously, templates like {{NUMBEROFARTICLES}} are doing this locally. Is an external pull possible? Can someone point me to documentation on how this might be done? Thanks, West.andrew.g (talk) 16:34, 3 December 2010 (UTC)

    It's not possible. variables like {{NUMBEROFARTICLES}} are part of MediaWiki or one of its extensions. Graham87 03:24, 4 December 2010 (UTC)
    In your scenario it sounds suitable to just make a template with the number, edit it periodically and say "as of (time)". Do you really need a live number? PrimeHunter (talk) 21:37, 4 December 2010 (UTC)
    Thanks everyone for your help. No, a live number isn't at all required -- I was just as interested in whether or not it was technically possible. Alternative to this "pull" approach, I could always write a 'cron' job to "push" the change, by using the MediaWiki API. Thanks, West.andrew.g (talk) 04:16, 6 December 2010 (UTC)

    Stricter semi-protection availability

    Over the past year I have been victim to a large number of vandalistic and extremely harsh edits on my talk page. These edits make fun of my spinal condition and general state of ill health. Several administrators have come in and semi-protected my pages for a period of time, but the vandal simply waits to be autoconfirmed (10 edits and four days) and continues this pattern of harassment. If I knew the real-world identity of this editor I would likely contact the police. Here is what I propose: Autoconfirmed users coming from Tor networks have stricter autoconfirmed threshold of 90 days and 100 edits. I ask that a new level of semi-protection be created that would restrict editing to those who have this type of history. Call it Semi Protection Level 2 or whatever. I don't just ask but I beg for this to happen. It's truly weighing very heavily on me to have my user and talk pages so frequently vandalized and in such a traumatic manner. Thank you. Basket of Puppies 01:56, 5 December 2010 (UTC)

    Users cannot edit from Tor nodes without exemption, per policy. Hence both TorBlock and TorNodeBot. --Shirik (Questions or Comments?) 02:00, 5 December 2010 (UTC)
    I realize, Shirik. I am asking that a new level of semiprotection be made that won't let someone edit until they have 90 days and 100 edits. This is separate from Tor. Basket of Puppies 02:02, 5 December 2010 (UTC)
    I think the point Shirik is making is that since people can't edit from Tor anyway, there wouldn't be much of a point in creating an additional level of semiprotection that would restrict people that use Tor. Ucucha 03:26, 5 December 2010 (UTC)
    Sorry BoP for having to go through so much, it's vandals like those that want to make me use a particular editor's "scare a vandal for life" technique. I have also wanted protection on the occasion, also from other established editors, such as userpage protection so that only I could edit it. Could it be implemented so that the owner of the userpage may set criteria for their user namespace, for who can edit given variables such as edit_count, account_age and user_name. My page, while it would blank if someone other than I edited it, would only take a little know how to change say, my PGP key. Being able to customise who can edit will give great security. Naturally, it would inherit regular page protection, (and not be able to stop admins from editing?) 930913 (Congratulate/Complaints) 02:34, 5 December 2010 (UTC)
    Technically, you can restrict pages to being edited only by yourself (and admins) by using .js or .css pages. For example, you could put your userpage at Special:Mypage/userpage.css, and have your actual userpage (Special:Mypage) just be {{:User:A930913/userpage.css}} and fully protected. Ucucha 03:24, 5 December 2010 (UTC)
    Right, but this doesn't work for me. I want editors in good standing to be able to communicate with me. I simply want them to be on the wiki for 90 days and have 100 edits. Please. Basket of Puppies 03:33, 5 December 2010 (UTC)
    How about creating (or adding to) an abuse filter? That delivers a more targeted solution to this long-term abuse without affecting the protection system. Netalarmtalk 03:53, 5 December 2010 (UTC)
    6 of one, half-dozen of another. I'll let the developers decide which is easier to implement, but I think the community consensus is important to gauge. Basket of Puppies 04:15, 5 December 2010 (UTC)
    But that still doesn't stop someone from changing it from {{User:A930913/up.css}} to {{User:random/lol}} 930913 (Congratulate/Complaints) 09:27, 5 December 2010 (UTC)
    It does, if your actual userpage is fully protected. Ucucha 13:41, 5 December 2010 (UTC)
    With full protection it disabled editors-in-good-standing from communicating with me. With a new level of semi-protection it would only enable editors with 90 days and 100 edits to do so. Basket of Puppies 17:49, 5 December 2010 (UTC)

    There may be other ways to deal with Basket of Puppies' specific problem, but I can see merit in having another tool at our disposal to handle vandals who learn how to game our system. An intermediate level of protection between allowing any account with 10 edits and allowing only admins could be very handy in special situations, tho it should be used very sparingly. --agr (talk) 02:34, 6 December 2010 (UTC)

    Where are these capital letters coming from?

    I've just edit a user's talk page, see User talk: OneDalm0#Alleged Roman head, where I now see "Maybe You Could Also look for sources and Stock pointing out the time frame That Does not make it totally impossible for this to Have Been" although I wrote, and you can see that when you click edit, "Maybe you could also look for any sources pointing out that the time frame doesn't make it totally impossible for this to have been sourced ". I see it in Chrome v.8, not in Firefox or Opera. Dougweller (talk) 06:37, 5 December 2010 (UTC)

    I don't see it too. I suspect it is something to do with Chrome settings... Rehman 06:39, 5 December 2010 (UTC)
    There's a translation service (Spanish to English, even though only part of the page is Spanish) in Chrome that introduces those caps. --MASEM (t) 06:41, 5 December 2010 (UTC)
    That makes sense. Thanks. Dougweller (talk) 07:06, 5 December 2010 (UTC)

    Special:Whatlinkshere

    Does anyone know if there is a way to hide links from templates without physically editing the templates themselves when looking at what links to any article? This way it would be possible to see what links there from an article or page itself and not from navigation. Simply south (talk) 16:33, 5 December 2010 (UTC)

    It's not possible with the special page, as far as I am aware. Ucucha 16:44, 5 December 2010 (UTC)
    You can filter out the templates themselves by selecting the 'Article' namespace, but it wil not hide the links the those templates produce on other articles. — Edokter • Talk • 16:52, 5 December 2010 (UTC)
    Could it be done? (Original query) Simply south (talk) 16:57, 5 December 2010 (UTC)
    I'm afraid not, unless a "Hide transcluded links" option were built into MediaWiki. — Edokter • Talk • 17:14, 5 December 2010 (UTC)
    That would certainly be helpful. Rd232 talk 17:23, 5 December 2010 (UTC)
    Do you think someone would do that? Simply south (talk)
    We can't do it. You would have to request that feature on Bugzilla. — Edokter • Talk • 17:30, 5 December 2010 (UTC)
    Perhaps someone could make that Bugzilla request (see also Help:Searching section up the page). Rd232 talk 17:47, 5 December 2010 (UTC)
    How do we get someone to write the report? I'm not a programmer. Simply south (talk) 23:36, 5 December 2010 (UTC)
    See bugzilla:3241 and an earlier discussion at Wikipedia:Village pump (technical)/Archive 69#What links here and navboxes. PrimeHunter (talk) 02:26, 6 December 2010 (UTC)

    New str templates

    I was trying to write new string templates to count preceding numbers and letters in a given string. But it looks like I exceeded the functions' limits. I have tried for the past two days to just undo my edits, and revert to the last good versions, but I keep getting an "Our servers are currently experiencing a technical problem" error. If anyone would like to give it a try, please undo my 4 December edits to Template:Str number and Template:Str letter. Thanks, 117Avenue (talk) 03:48, 6 December 2010 (UTC)

    Please please PLEASE go add a vote to bugzilla:26092 so we can just get rid of these ridiculous hacks already. I was unable to revert either template myself (I couldn't get {{Str letter}} to load at all), so I deleted {{Str number}} and restored the pre-December 4 revisions. 「ダイノガイ千？！」? · Talk⇒Dinoguy1000 04:26, 6 December 2010 (UTC)
    Thanks for fixing the one. You can try this backdoor link to get to the other one. 117Avenue (talk) 04:43, 6 December 2010 (UTC)
      Done. Can't believe I didn't think of that... If you still need the contents of the deleted revisions for some reason, just ask, but I'd recommend against trying to re-save them as live code. ;) 「ダイノガイ千？！」? · Talk⇒Dinoguy1000 06:09, 6 December 2010 (UTC)
    Thanks again. But I was hoping to keep the code, since it took a while to write. I was thinking of putting it on a page that wouldn't affect others (I think part of the problem was having documentation), and playing around with it to see if I could make it work. Do you have any other ideas to make it better? Currently, because of the excessive use of #if's, it only works up to 25 characters, and if transcluded, even less. I was hoping for at least 80. 117Avenue (talk) 07:47, 6 December 2010 (UTC)
    A better solution is to re-evaluate the problem you are trying to solve. What is your intended use for a "number of digits at the start of a string" function? OrangeDog (τ • ε) 12:40, 6 December 2010 (UTC)
    The only use I have applied it to is Template:Infobox election/sandbox, where it extracts the specified width of an image, if width and height have been specified. For example: {{Str number/trim|200x500px}} will produce 200. I imagine there could be other uses, which is why I created a corresponding number of letters template. 117Avenue (talk) 21:29, 6 December 2010 (UTC)
    Please, if it's possible, try to avoid using string templates in their current form. It seems you have some image template and want to accept parameters like size=200x500px, right? In that case, I think it would be better if you used two parameters, e.g. width=200|height=500. That way, you don't have to use string functions at all. Svick (talk) 21:57, 6 December 2010 (UTC)
    It is a widely used template, I can't do a parameter change on my own. I haven't come across any cases where map_size isn't in the 200px format, but I want to be able to use this parameter in setting the box width. 117Avenue (talk) 00:32, 7 December 2010 (UTC)
    You can use a bot to rewrite the pages to use parameters of the proper format. Using elaborate string functions is NOT the method that you should use to fix this problem. —TheDJ (talk • contribs) 22:22, 7 December 2010 (UTC)

    Editor says IP is blocked, looks unblocked to me

    Ottsby (talk · contribs) was blocked this morning for recreating an article that was thought (wrongly) to be a hoax. He's been unblocked but says 69.143.12.8 (talk · contribs) is still blocked. It shouldn't be, I've checked and it doesn't show up as blocked, and I thought I'd lifted the autoblock this am. Why is he still getting a block message when he tries to edit from his account? Thanks. Dougweller (talk) 17:47, 6 December 2010 (UTC)

    http://toolserver.org/~nakon/autoblockfinder.php?u=Ottsby ΔT The only constant 18:22, 6 December 2010 (UTC)
    Weird, I did that this morning. I didn't think to check. Thanks. Dougweller (talk) 18:39, 6 December 2010 (UTC)

    Database download down?

    When will http://download.wikimedia.org/ be back?Smallman12q (talk) 01:50, 7 December 2010 (UTC)

    See what changes at wikitech:Dataset1#11-10-2010 - New errors. Until then...? – Allen4names 17:27, 7 December 2010 (UTC)
    Per wikimedia-tech IRC channel, the vendor is taking long to respond (and fix), and the guy dealing with those systems is a volunteer, so it can take some additional time in communication and execution there as well. The hope is that this week it will become known what the ETA with the vendor will be. —TheDJ (talk • contribs) 22:39, 7 December 2010 (UTC)
    OK, apparently the problem is that the parts require firmware updates that can only be run from a server with Windows installed, which this server OF COURSE does not have. Updates will be posted here. —TheDJ (talk • contribs) 22:45, 7 December 2010 (UTC)

    Google caching not picking up BOT vandalism reverts

    Various flavors of this topic have been discussed a handful of times in the past, but I wanted to resurrect the conversation with a specific new twist: Bot reverts are not triggering the same refresh in google that regular edits do.

    An OTRS ticket pointed me to the issue in relation to this vandalism. ClueBot NG reverted the vandalism in the same minute, but none the less google picked up and maintained the vandalism for a number of hours. I followed procedure and logged into google webmaster tools and requested removal, but it was denied as google claims that the page is not available in their cache (which is incorrect). So I made a few cleanup edits and within 30 seconds the cached and preview version on Google had been updated.

    This is leading me to believe that there is something unique that we have (perhaps intentionally) done in the feed to filter out bot edits. Any thoughts?  7  06:18, 7 December 2010 (UTC)

    Yeah, I can understand how that could be. The bot is maybe looking only at the diff page or through the API rather then look at the page as a human browsing would do. Regards, SunCreator (talk)
    I don't think Google would be able to see that. What Google is probably doing is ignoring edits by bots (or using a recent changes feed that excludes bots). That makes sense for things like interwiki bots, but not for ClueBot. Ucucha 13:12, 7 December 2010 (UTC)
    Google sees. No reason to exclude bots unless Google told somewhere i.e robots.txt or noindex on page. Regards, SunCreator (talk) 00:30, 8 December 2010 (UTC)
    I think we are all agreeing that there is no reason for google to exclude bot changes in its refresh, the question now (perhaps for the developers) is whether this is something that needs to change on WP side or on google's side.  7  00:37, 8 December 2010 (UTC)
    I think there is a difference but I don't know how the reverting bots work exactly so I can't be sure. I think the revert processes are generally as follows:
    1. (Bot process) From recentchanges, open a diff page, hitting undo and save, goto next recentchanges item
    2. (Human process) Browse to view an article page, go into history, hitting undo and then save and then view the amended page
    The above two processes are different, first process is by a bot, while second is always human. In the first process the actual article is not viewed directly - either before or after. In the the second the article is viewed twice, once before and once after the change, Google can pick up on this viewing and give it be reason to know the page contents has changed and change it's copy accordingly. Regards, SunCreator (talk) 01:02, 8 December 2010 (UTC)
    How would Google know that someone views a page on Wikipedia? Ucucha 01:07, 8 December 2010 (UTC)
    Google gets the information from ISP's who are serving the page to the end user. Regards, SunCreator (talk) 01:13, 8 December 2010 (UTC)
    (edit conflict)My understanding is that Google sent the Foundation a server that just sits in Florida and runs Googlebot really fast over every changed page. Even though there's no way for this setup to detect directly whether a page has been viewed, what Google sees will probably be the last version of the page that was saved in the SQUID cache. Logged-in users (such as all approved bots) never view pages from the cache, so they might not cause the page cache to be refreshed when they view a page, but logged-out users always view pages from the cache and can cause the cache to get refreshed. If a vandal on an IP causes the page cahce to be refreshed, and the bot revert immediately afterward doesn't, that would account for the observed behavior, especially as CluebotNG is now responsible for such a large percentage of vandalism reverts. Assuming I'm right, there's a reasonable fix: if CluebotNG explicitly purges pages after it reverts, that would assure that its reverts were the most recent version in the cache, at the cost of stressing the cache servers a bit. On the other hand, I'm not a server admin, so I can't guarantee that this would fix the problem. — Gavia immer (talk) 01:24, 8 December 2010 (UTC)
    Interesting. Could even be a hybrid of the two situations, external knowledge of user viewing article causing internal googlebot to re-read page off Florida server. Whatever is happening it sounds like there is reason to believe that a bot revert and a human edit is not going to have the same effect on Google getting the latest copy of an article. Regards, SunCreator (talk) 01:40, 8 December 2010 (UTC)
    Could this be due to something as simple as the fact that CluebotNG marks its reverts with the minor flag? Conceivably GoogleBot might presume that timely google DB updates aren't needed for "minor" edits. -R. S. Shaw (talk) 03:53, 8 December 2010 (UTC)
    According to Cobi and Tim Starling is likely a conincidence. Cluebot apparently doesn't use the Bot flag or the minor flag. I was also able to see a cluebot revert in the recent changes RSS feed... so it doesn't seem like it's anything under our control. Nothing left to do I guess.  7  04:32, 8 December 2010 (UTC)

    Article won't load

    The article Football League Trophy that I've previously read/edited is not currently loading. Anyone else having the same issue, and if so, any ideas on the problem? Thanks. Eldumpo (talk) 12:59, 7 December 2010 (UTC)

    Loaded fine just now. Do you get an error message or no response. Something from server slow, proxy error threads above perhaps? Regards, SunCreator (talk) 13:14, 7 December 2010 (UTC)
    No problem for me either. What exactly is the problem you encounter? Ucucha 13:32, 7 December 2010 (UTC)
    It was just going completely blank, no Wiki logo or anything, but anyway it's working fine for me as well now. Always the way! Thanks. Eldumpo (talk) 13:39, 7 December 2010 (UTC)

    Broken userpage

    My userpage is broken. Scroll down in the userboxes, and my stuff is there. Can you fix that? Thanks! --Perseus (t • c) 17:57, 7 December 2010 (UTC)

    Depends what you are trying to do. I've made a change that hopefully is what you wanted. Rambo's Revenge (talk) 18:01, 7 December 2010 (UTC)
    I want my userboxes on the right, separate from everything else. Perseus (t • c) 18:03, 7 December 2010 (UTC)
    Looks like this right now. Perseus (t • c) 18:13, 7 December 2010 (UTC)
    It's   Not done yet. Perseus (t • c) 18:49, 7 December 2010 (UTC)
    One, you need to wait more than 36 minutes before complaining that your userpage hasn't been fixed. Secondly, I honestly don't know what you're trying to do; it's obvious that you've thrown together a bunch of HTML code that you don't actually know how to use. Perhaps start from scratch? (and don't transclude the userboxes until you've got it looking the way you do; part of the problem is that I can't tell which HTML code needs to be fixed) EVula // talk // ☯ // 21:25, 7 December 2010 (UTC)

    When I try to edit the "references" section on this link, the box that comes up is completely blank!

    Hi . . . I'm a pretty experienced contributor and have never run into this issue before. I tried to edit the "references" section on the following existing article:

    Sandy and Beaver Canal

    Though five references are listed when you simply go to the article, when you click "edit" here, the box that appears onscreen is empty! So I'm spooked about editing it because I don't want to accidentally delete something that's already there.

    All I want to do is add information about The Sandy & Beaver Canal, a book co-written by william H. Vodrey III & R. Max Gard. It's been out of print for ages, but is available on CD-ROM and can easily be found on Ebay.

    Ideas?!? Thank you! —Preceding unsigned comment added by Beaverkettle (talk • contribs) 18:18, 7 December 2010 (UTC)

    Just {{reflist}} Perseus (t • c) 18:23, 7 December 2010 (UTC)

    You need to add the in-text citation within <ref>...</ref> tags. See Wikipedia:Referencing for beginners. ---— Gadget850 (Ed) talk 18:26, 7 December 2010 (UTC)
    (e/c) It's not empty, it consists of {{reflist}}, which inserts formatted references given in <ref> tags appearing above in the article text. This is quite standard. You can safely insert other references below the reflist, though then it might be better to split the section to "Notes" and "References".—Emil J. 18:27, 7 December 2010 (UTC)

    Weird behaviour when trying to transclude "special:" pages

    I have stumbled onto this accidentally, so this is again something more of a "completely useless curiosity". A user (Silenzio76) tried to transclude the "Special:Recentchanges" page on his user page using {{Special:Recentchanges/100}}. It was later removed. Here is the version without it, and here is the version with the transclusion (this is the diff that removes it). Now, with {{Special:Recentchanges/100}} removed the page looks and reads perfectly fine. But when that code is used, several strange things happen:

    • The title of the page changes to Special:Recentchanges/100.
    • More mysteriously, the two sections on the page acquire weird strings, starting with "UNIQ" and ending with "QINU", with a random sequence of characters/numbers between, e.g. UNIQ151b7add547bbd-h-0--QINU and UNIQ151b7add547bbd-h-1--QINU (with some framing unicode characters which I am unable to reproduce).

    I guess transcluding "special:" pages is something of a Bad Thing(TM)... shouldn't it perhaps be entirely disallowed, rather than leading to such bizarre results? -- 80.135.1.231 (talk) 20:43, 7 December 2010 (UTC)

    I think this bug has already been reported at Bugzilla somewhere—can't find it back though. Ucucha 20:52, 7 December 2010 (UTC)
    Found it: bugzilla:23293. Ucucha 20:56, 7 December 2010 (UTC)
    It is marked as fixed though, and was so before the changes that OP notet. Taemyr (talk) 21:06, 7 December 2010 (UTC)
    The fix may not have been deployed yet on the version of MediaWiki that Wikipedia is running. Ucucha 21:11, 7 December 2010 (UTC)
    Thanks everyone -- according to the Bugzilla page "Wikimedia sites are running a version of mediawiki before the version where this is fixed." (2010-10-07) -- this explains why the problem is still present. I guess this will be fixed when we upgrade to the newest version of MediaWiki. -- 80.135.1.231 (talk) 21:16, 7 December 2010 (UTC)
    More precisely, it will be fixed when r69820 gets reviewed. Ucucha 21:20, 7 December 2010 (UTC)
    Not quite. Unless someone specially applies that revision to the site, it could sit as reviewed for quite some time until the next "regular" update. IIRC, the last "regular" update was in April 2010; those in charge of such things are hoping to do better once they catch up with the backlog. Anomie⚔ 21:39, 7 December 2010 (UTC)

    Template loop warning

    I got a "template loop warning" at this current AfD page: Salesian High School (Richmond, California)Cullen328 (talk) 04:57, 8 December 2010 (UTC)

    Mostly because it was calling to transclude itself. I've fixed the page. If you'd like to fix the rest (explanation of who started the AFD, etc), feel free. Killiondude (talk) 05:07, 8 December 2010 (UTC)
    This page still doesn't appear to be properly transcluded: the comments at Wikipedia:Articles for deletion/Log/2010 December 8#Salesian High School (Richmond, California) are not showing up at Wikipedia:Articles for deletion/Salesian High School (Richmond, California). This is beyond my meager understanding of how the AfD templates and transclusions work. --Arxiloxos (talk) 05:40, 8 December 2010 (UTC)
    I see what happened. I think I've fixed it. I'm not too familiar with how AFD is formed (I usually just use Twinkle), so I tried to mimic what other AFD discussions are doing. Killiondude (talk) 05:47, 8 December 2010 (UTC)
    Looks like it's fixed now--bravo, and thanks, Killiondude.--Arxiloxos (talk) 05:49, 8 December 2010 (UTC)

    Secure server slow

    The secure server https://secure.wikimedia.org, is taking a long time to load Wikipedia pages, about 10-30 seconds to load a page and often fails to save pages, meaning that you retry often - and less included inclined to preview. Anyone else having the same issue? Regards, SunCreator (talk) 18:07, 4 December 2010 (UTC)

    Fine for me in London... ╟─TreasuryTag►You may go away now.─╢ 18:09, 4 December 2010 (UTC)
    It's has still been going slow but noticed it waits a long long time on images. I recently switched off images on my browser and all is nice and fast now. My firewall has logging many errors of traffic to unopen ports coming from secure.wikimedia.org, so perhaps that is mis-configured or something. Regards, SunCreator (talk) 23:34, 4 December 2010 (UTC)
    Hm, fine for me on my end, I recently got a new internet service. - Dwayne was here! ♫ 03:06, 5 December 2010 (UTC)
    Was horribly slow for me in Sri Lanka a few months back. But it's fast enough now. Regards. Rehman 03:21, 5 December 2010 (UTC)
    I get about 50% failure rate on reading pages. The page never comes up and I have to reload and try again. Regards, SunCreator (talk) 01:49, 9 December 2010 (UTC)

    502 proxy error

    Received the following a moment ago when hitting 'save page'

    502 proxy error

    Proxy Error

    The proxy server received an invalid response from an upstream server. The proxy server could not handle the request POST /wikipedia/en/w/index.php.

    Reason: Error reading from remote server

    Apache/2.2.8 (Ubuntu) mod_fastcgi/2.4.6 PHP/5.2.4-2ubuntu5.7wm1 with Suhosin-Patch mod_ssl/2.2.8 OpenSSL/0.9.8g Server at secure.wikimedia.org Port 443

    Regards, SunCreator (talk) 18:00, 6 December 2010 (UTC)

    I have also seen this error a few times, and got it again today, with the same error message SunCreator got. Ucucha 15:46, 7 December 2010 (UTC)

    The secure server is almost continuously at maximum load. The service will be moved soon (to the new datacenter in virginia), and the hope is that this should be a less frequent occurrence. —TheDJ (talk • contribs) 22:42, 7 December 2010 (UTC)
    Thank you. Regards, SunCreator (talk) 00:25, 8 December 2010 (UTC)

    Bug page on Bugzilla --vgmddg (look | talk | do) 23:25, 8 December 2010 (UTC)

    No orange bar

    I've received three new messages over the past 24 hours, but not a single orange bar. Could someone please look into this? --Dylan620 (t • c • r) 01:17, 7 December 2010 (UTC)

    Works for me... Rehman 13:47, 7 December 2010 (UTC)
    Make sure you aren't using some custom css or js that could be preventing the banner from showing. I think there is an option in user preferences for enabling/disabling this, so check in to that as well. - ʄɭoʏɗiaɲ τ ¢ 20:29, 8 December 2010 (UTC)

    Msnbot

    Out of a completely useless curiosity, why does the toolserver disallow msnbot [26]? Magog the Ogre (talk) 19:21, 7 December 2010 (UTC)

    http://journal.toolserver.org/entry/2009/06/09/wherein-msnbot-behaves-badly-and-is-banished/ — Gavia immer (talk) 19:27, 7 December 2010 (UTC)
    That article seems confusing. If the 3.5 million articles(and other pages) from Wikipedia are on toolserve then a bot looking at the original source would need to be reading at those rates per second otherwise it will soon be out of date. Regards, SunCreator (talk) 19:48, 7 December 2010 (UTC)
    This isn't the main wiki servers we're talking about, and they don't serve a public copy of the encyclopedia; see meta:Toolserver for the sort of things that are hosted there. — Gavia immer (talk) 20:00, 7 December 2010 (UTC)
    Okay, how many pages would meta:Toolserver have? Less then a million I take it. Regards, SunCreator (talk) 00:26, 8 December 2010 (UTC)
    Read the link I gave again - msnbot was supplying random query parameters to the tools hosted on the toolserver. It doesn't take all that many permutations of such parameters for the total number of page requests to grow very large. — Gavia immer (talk) 00:36, 8 December 2010 (UTC)
    Depends if those parameter where logically crawl-able pages. i.e pages that humans could be expected to visit, if they are then bots crawl them, if there not then how did msnbot know about them? Regards, SunCreator (talk) 02:43, 8 December 2010 (UTC)
    I don't know anything that isn't in the page I linked, and I doubt that anyone else here does either. If you want to know something that specific, I'd take it up with the toolserver administrators. — Gavia immer (talk) 02:48, 8 December 2010 (UTC)

    Random parameters? They're liable to find a security hole and bring down the whole server. Very interesting that the bot would do something so obnoxious. And not surprising at all that MS was too big of a bureaucratic nightmare to do anything about it. Magog the Ogre (talk) 23:12, 8 December 2010 (UTC)

    Javascript help

    Anyone know why the div I insert on a diff page in this code doesn't work? If I set it to a specific px value (i.e. "width:500px"), the div works. If, however, I set it to a percentage (i.e. "width:100%"), it doesn't. Maybe I'm just missing something?--Dudemanfellabra (talk) 01:42, 8 December 2010 (UTC)

    Nvm I did it a different way. Still would like to know what caused that width discrepancy, though...--Dudemanfellabra (talk) 00:37, 9 December 2010 (UTC)

    Need a template to change color of section

    At Talk:United States diplomatic cables leak#Removed content I added a copy of a mainspace section using <blockquote>. Is there some code or existing template wrapper that I could also employ to give this text a ligher or different background color to clearly set it apart from the surrounding discussion? __meco (talk) 16:22, 8 December 2010 (UTC)

    <blockquote> accepts style attributes:

    Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

    Algebraist 16:26, 8 December 2010 (UTC)
    Great wonder! __meco (talk) 16:36, 8 December 2010 (UTC)

    ParserFunction errors

    Can someone have a look at Odontogriphus. The first clue that there is a problem, is the red error at the bottom of the page. However, this is not the core issue. If you look in the references, you will see loads of errors before that. Essentially, it appears that some parserfunction expansion limits are being broken, which is causing errors downstream. To help track down the cause, you can do a "show preview" on just the lower section, to see that the last red error message is not the issue. You can further copy sections of the article into a sandbox to see what is causes the error. I was not able to figure it out with my limited tests. Thank you! 134.253.26.10 (talk) 16:54, 8 December 2010 (UTC)

    Update, I think it might have something to do with the taxobox or automatic taxobox. 134.253.26.10 (talk) 17:04, 8 December 2010 (UTC)
    I don't know why, but moving that section above the references section seemed to solve the problem. — Martin (MSGJ · talk) 17:08, 8 December 2010 (UTC)
    No, it didn't solve the problem. Read the references, loads of "WARNING: argument omitted, expansion size too large" errors. :( 134.253.26.10 (talk) 17:12, 8 December 2010 (UTC)
    Fixed by removing the automatic taxobox, which is probably using so many templates that the parser doesn't like it. Might the fact that you called Lophotrochozoa subphylum instead of a superphylum have confused something? Ucucha 17:16, 8 December 2010 (UTC)
    I don't know, I didn't fill out either taxobox. I was just cleaning up Category:ParserFunction errors. Thank you for fixing it. Hopefully the taxobox folks will be able to fix the "automatic taxobox". 134.253.26.10 (talk) 17:20, 8 December 2010 (UTC)

    Trouble at WP:VPA

    Currently the Technical Village Pump isn't showing up at all at WP:VPA. I don't know enough to figure out the problem. Ntsimp (talk) 21:27, 8 December 2010 (UTC)

    Fixed. Algebraist 21:31, 8 December 2010 (UTC)

    consolidating logged-out edits

    Is it possible to consolidate logged-out edits into my account record? It's happened a couple times, but the main IP was this one. (No-one else has ever used it.) Not just for edit-countitis, but also for anonymity. — kwami (talk) 21:32, 8 December 2010 (UTC)

    No. You can rev-delete away the IP number, though. Ucucha 21:34, 8 December 2010 (UTC)
    Not any more. :-) Graham87 02:06, 9 December 2010 (UTC)

    2 sections in 2 side by side columns

    Is it possible to put see also and external links (or any other 2 sections or subsections) in 2 side by side columns.
    Just granpa (talk) 23:02, 8 December 2010 (UTC)

    Nevermind. Template:Multicol seems to be working fine.
    Just granpa (talk) 00:44, 9 December 2010 (UTC)

    Higher protection level

    Could a higher protection level be applied to certain pages that would only be overridden by a bureaucrat and would prevent admins from performing any actions regarding such a page? Creating this new level can ideally prevent wheel wars such as this and edit wars like this from occurring between sysops on a particular bureaucrat-protected page. :| TelCoNaSpVe :| 06:00, 9 December 2010 (UTC)

    Hmm, something interesting. But I wonder if the amount of work required to do this is worth considering the number of pages that require this type of protection... Rehman 06:36, 9 December 2010 (UTC)

    Long-time donator and really annoyed by the banner

    Hello,

    I've donated to the Wikimedia Foundation every year since I became an active contributor and user (and thus an utlilizer of the Foundation's bandwidth and resources).

    So this year I was again proactive and donated as soon as the donation campaign begun (I was actually looking forward to it, as I know the fundraising run takes place around December, when we anyway think about giving gifts). I closed the banner after I had donated. I always donate generously to the Foundation, as Wikipedia is absolutely invaluable to me personally.

    Since then, I have had to close the donation banner really multiple times, and it is really starting to annoy. Every time I close the banner, it keeps reappearing after a seemingly random period of time, even if I am logged-in and the system surely knows that I have already donated / clicked the close icon. The random time period after it bugs me again is a telltale sign that this behaviour is purposeful, something akin to the psychological irritation tactics that advertisers use against the public in TV commercials to force their message on them.

    I don't think the Wikimedia Foundation needs this kind of tactics; its success is based on the users trusting and appreciating the Foundation, and aggressive fundraising tactics can turn the public's opinion, especially as the Foundation's current finance is fine: it has over $12.7M in reserves, more than all Foundation expenses combined between 6/09 and 6/10.

    I read some other dissatisfied comments to learn about the gadget that hides the banner. This is, however, not really a nice solution for multiple reasons:

    • I won't be able to see the next years banner (same goes to AdBlocking it)
    • The solution requires work and possibly luck to find
    • It is counterintuitive that closing the banner does so only for a random period of time, and then it will reappear again on every single page

    Can you please make it so that logged-in users can just hide this year's banner simply by closing it? I don't think we need to be subjected to confrontational / aggressive advertisement just because we contribute to a free online encyclopedia, or?

    Also, the biggest sufferers of the status quo are of course the anonymous users, who don't want to, or haven't had time to, create an account, and thus have no way to close the fundraiser banner except temporarily. We might be bugging off valuable future contributors by giving a wrong image of the values Wikipedia stands for and how it treats its users. --hydrox (talk) 05:59, 5 December 2010 (UTC)

    • I have agreed to the same topic before, and I will again. I've been an active editor for quite some time now. And there's no word to say how annoying it is (even though there is an option to suppress it at Special:Preferences). Rehman 06:17, 5 December 2010 (UTC)
    If the code is working as written, the close button will shrink the banner for 7 days. Is your complaint that 7 days is too short, or is it reappearing more often than that? Dragons flight (talk) 05:45, 6 December 2010 (UTC)
    Sometimes it appears once a day, sometimes it'll appear again on the next page after I "X" it. Rarely, if ever, has it remained hidden for a full seven days. I feel the donation I make is in the myriad of content which I add to the encyclopedia; I have time, I don't have money. I'm tempted to write a piece of css code that will permanently hide it from my sight. - ʄɭoʏɗiaɲ τ ¢ 20:22, 8 December 2010 (UTC)
    What I find even more annoying is that not all projects have an option to suppress these banners -- or at least not one I can find under "Preferences". (And I have looked. Carefully & often.) What I am forced to do is to stop contributing to -- & viewing -- those projects until the fundraiser is over. Some of us volunteers believe improving the content is a sufficient donation to Wikimedia projects that we shouldn't be hounded for money, too. (I found having the box to click to make these banners go away was a very nice thing -- while those boxes worked. I can only guess that the reason these boxes are no longer part of the banners is that donations weren't generous enough.) -- llywrch (talk) 19:51, 9 December 2010 (UTC)
    You can still hide the banner on any wiki by adding "#centralNotice{display:none;}" to your .css page. Ucucha 20:01, 9 December 2010 (UTC)
    Oh well. Bye bye fundraising banners, you did it to yourself. What is the link to the discussion on the implementation of this banner? Where is the community consensus to have these pop up at random? - ʄɭoʏɗiaɲ τ ¢ 00:17, 10 December 2010 (UTC)
    1. ^ Refers to the Maine Coastal Island Registry ("MCIR"), a comprehensive list maintained by the state conservation agency assigning each island a number, since many islands have the same names. Bar Island in Somes Sound for instance is MCIR #59-265, while Bar Island off the town of Bar Harbor is MCIR #59-194.
    Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia:Village_pump_(technical)/Archive_82&oldid=1208876166"