User talk:Σ/Archive/2014/April

WikiCup 2014 March newsletter

A quick update as we are half way through round two of this year's competition. WikiCup newcomer Godot13 (submissions) (Pool E) leads, having produced a massive set of featured pictures for Silver certificate (United States), an article also brought to featured list status. Former finalist Adam Cuerden (submissions) (Pool G) is in second, which he owes mostly to his work with historical images, including a number of images from Urania's Mirror, an article also brought to good status. 2010 champion (Pool C) is third overall, thanks to contributions relating to naval history, including the newly featured Japanese battleship Nagato. Cliftonian (submissions), who currently leads Pool A and is sixth overall, takes the title for the highest scoring individual article of the competition so far, with the top importance featured article Ian Smith.

With 26 people having already scored over 100 points, it is likely that well over 100 points will be needed to secure a place in round 3. Recent years have required 123 (2013), 65 (2012), 41 (2011) and 100 (2010). Remember that only 64 will progress to round 3 at the end of April. Invitations for collaborative writing efforts or any other discussion of potentially interesting work is always welcome on the WikiCup talk page; if two or more WikiCup competitors have done significant work on an article, all can claim points equally. If you are concerned that your nomination—whether it is at good article candidates, a featured process, or anywhere else—will not receive the necessary reviews, please list it on Wikipedia:WikiCup/Reviews. If you want to help out with the WikiCup, please do your bit to help keep down the review backlogs! Questions are welcome on Wikipedia talk:WikiCup, and the judges are reachable on their talk pages or by email. Good luck! If you wish to start or stop receiving this newsletter, please feel free to add or remove yourself from Wikipedia:WikiCup/Newsletter/Send. J Milburn (talk • email), The ed17 (talk • email) and Miyagawa (talk • email) 22:55, 31 March 2014 (UTC)

Archiving minor?

Hello.

I see that the Lowercase sigmabot III is marking archive edits with mb. It sure is a bot edit, but is archiving a section really a minor one?

HandsomeFella (talk) 21:09, 4 April 2014 (UTC)

(talk page stalker) Yes, see Wikipedia:Minor#Exceptions NE Ent 10:19, 7 April 2014 (UTC)

Bug in Lowercase III?

Hi, the bot just archived a number of old sections from my talk page but left a part out making a bit of a mess. See the diff here. Scroll to about the middle of the page and you'll see that a small portion of a section that was originally collapsed was not archived. I fixed the archive manually moving the content left behind from here to here. Regards. Gaba ^(talk) 22:01, 5 April 2014 (UTC)

The bot archives sections. It distinguishes sections by the ==Heading== syntax, and doesn't look for any {{collapse top}} that may be present. If you want something that looks like a heading to appear inside a collapsed box, but not be treated as the start of a section, you should use {{fake heading}}. --Redrose64 (talk) 23:22, 5 April 2014 (UTC)

Ok I see. I didn't create those sections, an unexperienced editor did on my TP so I collapsed all that garbage. Didn't know Lowercase couldn't handle sections within a collapsed environment. Thank you. Gaba ^(talk) 01:14, 6 April 2014 (UTC)

(talk page stalker) Not a bot problem. The bot handles collapsed sections properly. These sections were not individually collapsed, nor were they sub-sections of another section which were all collapsed. What was done was the collapse templates were used to make one large collapse spanning multiple sections. The sections in question were all level 2 headers. There is nothing in the sections which indicates that they are connected. The sections left on the page by the bot do not have dates and were properly left on the page. The existence of any particular template (examples: {{collapse top}} {{collapse bottom}}) do not logically connect the sections in any way from the point of view of the MediaWiki software. If you had wanted the portions of the page which were collapsed to be considered all one section while collapsed then there should have been one level 2 header with level 3 headings (e.g. ===Comment being replaced===) under it.

I know that this might appear to be something which is counter-intuitive. However, it is those headings which define how the MediaWiki software handles portions of a page, not any particular template. — Makyen (talk) 02:15, 6 April 2014 (UTC)

Thank you for the detailed explanation Makyen, it sounds like a completely reasonable way to work. I presume this is a very rare case of the bot not doing what the user would have thought it would do. Regards. Gaba ^(talk) 03:02, 6 April 2014 (UTC)

Save us, sigmaboat!

WP:ANI is becoming an unarchived mess. We have this in the very top (as of ~ 06:20, 7 April 2014‎ ), but the bot didn't pick up on its run around 2145 the same day:

<noinclude>{{Wikipedia:Administrators' noticeboard/IncidentsHeader}}</noinclude>__NEWSECTIONLINK__
{{User:MiszaBot/config
|archiveheader = {{Administrators' noticeboard navbox all}}
|maxarchivesize = 700K
|counter = 826
|algo = old(36h)
|key = aad625193afdee54f00c742ee5ab61d1
|archive = Wikipedia:Administrators' noticeboard/IncidentArchive%(counter)d
}}
<!--
{{User:NOTClueBot III/ArchiveThis  <!--This used to be correct, but CB3 finds it in the page
and incorrectly archives -->
|header={{Administrators' noticeboard navbox all}}
|archiveprefix=Wikipedia:Administrators' noticeboard/IncidentArchive
|format=%%i
|age=36
|index=no
|numberstart=826
|archivenow={{tlu|User:ClueBot III/ArchiveNow}}
|minarchthreads= 1
|minkeepthreads= 4
|maxarchsize= 700000
|key=d85a96a0151d501b0ad3ba6060505c0c
}} -->
<!--
-----------------------------------------------------------
New entries go down at the *BOTTOM* of the page, not here. 
...

Any ideas what we're doing wrong? (The uncommented part used to work back when MiszaBot was running). NE Ent 09:06, 8 April 2014 (UTC)

Salvation! Thanks. NE Ent 09:26, 8 April 2014 (UTC)

(talk page stalker) @NE Ent: Does the "Salvation" comment mean this is resolved? Or is that just a comment that someone has manually archived the page?

lowercase sigmaBot 3 (lcsb3) begins its run at 0TC each day. The config template has been uncommented for one run, so some action should have been taken by the bot. This means that for any changes made we will not know if it worked for another 12 hours or so.

• Have you verified the |key=? Has lcsb3 run on the page ever? I know MiszaBot did; but in quickly looking, I don't see any prior action from lcSB3. NOTE: The key has changed. It was different when MiszaBot was operating, it was another different number prior to the recent counter change to 834, and it is now a third different number. The value of the |key= must be accurate or lcSB3 will not perform any archiving. The change to a new number implies that the "Salvation" comment above does indicate this situation is resolved.
• The |counter= was only just now changed from the old number of 326 to the current correct number of 334. I do not believe that having the wrong number would prevent the bot for archiving. However, threads possibly would have gone into the wrong archive page.

[NOTE: This 326->334 issue means that your recent use of OneClickArchiver resulted in the archives going into Wikipedia:Administrators' noticeboard/IncidentArchive826 instead of the correct archive of Wikipedia:Administrators' noticeboard/IncidentArchive834.) — Makyen (talk) 12:04, 8 April 2014 (UTC)

On what appears to be a related note, the archiving failed due to my internet connection being broken as it was archiving. Fortunately, it appears that SarekOfVulcan has noticed this and completed the archiving. →Σσς. (Sigma) 21:49, 8 April 2014 (UTC)

I changed the pre-formatted text so ClueBot does not erroneously archive the page. — Makyen (talk) 18:22, 9 April 2014 (UTC)

Leaves unwanted extra spaces in lead

In this edit, for instance, the bot has left unwanted extra space in the top of the article. Is there a fix to prevent that? Jason Quinn (talk) 18:17, 3 April 2014 (UTC)

I'll get to it one day, I promise. →Σσς. (Sigma) 07:11, 8 April 2014 (UTC)

https://tools.wmflabs.org/sigma/created.py

Don't work with startdate https://tools.wmflabs.org/sigma/created.py?name=Sunpriat&server=ruwiki&max=500&startdate=20140101&ns=&redirects=coalball Sunpriat (talk) 14:54, 6 April 2014 (UTC)

I've fixed it. Thanks for bringing it up. →Σσς. (Sigma) 07:11, 8 April 2014 (UTC)

Vertical whitespace?

See also: User talk:Misza13 § Protected edit request on 8 April 2014

Hello there! Why does including User:MiszaBot/config generate vertical whitespace on a page? It can be seen here, for example. Already checked the source of User:MiszaBot/config, and I'm not sure what's the cause. Could it be up to generating multiple subsequent line breaks? Would including a bare comment in the source help? Please advise. — Dsimic (talk | contribs) 01:11, 7 April 2014 (UTC)

(talk page stalker) As you probably noticed, {{User:MiszaBot/config}} consists of an {{#if:}} test directly followed by a <noinclude>...</noinclude>. If |key= is blank or absent, the if-test fails, and essentially what is left is one blank line. Since there is already a blank line immediately below, what you get between the {{WikiProject Linux}} banner and the first section heading is two blank lines. The TOC is automatically inserted before the first section heading, and so the gap before the TOC is the same as if the page had begun with

{{Talk header}}
{{WikiProject Linux|class=Template}}


= Not useful ==

As far as Wiki markup goes, altering the if test within {{User:MiszaBot/config}} so that it always emits something, e.g. a null HTML tag:

{{#if:{{{key|}}}|[[Category:Pages archived using a key]]|<span />}}

would fix it, but that might compromise the bot's actions. I'll let Sigma answer that one. --Redrose64 (talk) 08:58, 7 April 2014 (UTC)

(talk page stalker) It would be nice to fix this. However, it has been this way for years. The fix described is unlikely to affect the bot.

I have modified Template talk:Linux to add the text I typically use in addition to lcΣB3 configs (an archive box and the archive indexer bot config). Included is how I typically resolve not having a space between the header and the TOC. This is done by including __TOC__{{clear}} on the same line as the end of the {{User:HBC Archive Indexerbot/OptIn}}. The {{clear}} is to prevent the archive box from extending down into the first section when the TOC gets short. Example:

}}__TOC__{{clear}}

There are some other templates that are used in various headers which also output a blank line. In which case, a solution is to enclosing the newline separating templates within a wikitext comment which retains the readability of keeping the templates on separate lines. For example:

End of a template}}<!--
-->{{other template|

The configuration I added to Template talk:Linux also lines up the archive box vertically with the TOC. This is done by adding |style=margin-top:2px; to {{Archives}}. — Makyen (talk) 18:37, 7 April 2014 (UTC)

Thank you both for detailed insights, and for further improvements to Template talk:Linux. Totally appreciated!

The analysis pretty much aligns with my quick look into the bot configuration, which results in two subsequent line breaks leading to vertical whitespace. Would having an HTML comment instead of a <span /> element in "else" be a better option regarding the bot's functioning? A simple test confirms that such code also doesn't result in additional vertical whitespace, so something like this might be another option:

Some text.
<!-- config template resulted in no output -->

Some more text.

Makyen, your workaround works really great, thank you for that, but in my opinion that's just too much of a workaround and it should be instead properly fixed in templates. Having such code all around‍—‌no matter how usable it actually is‍—‌would be just too much and pretty much unreasonable. It's the 21st century and we have powerful computers, for God's sake. :) Hope you agree, and let's see what Sigma says. — Dsimic (talk | contribs) 03:33, 8 April 2014 (UTC)

I agree that it would be nice if it were fixed. I developed the workaround at a point when it looked like there wasn't anyone around to "fix" it. The workaround is part of one of my permanent clipsets which I drop onto a page when I want to configure archiving and just change the age parameters. I no longer think about it (much). Part of the problem was that the template has been incorporated into the use of lcΣB3, but is still under the MiszaBot user page. If you want a change in that template, this conversation should move to User talk:MiszaBot/config (it is redirected), which is the talk page for that template.

Changing what the template outputs should have no affect upon lcΣB3. From brief looks at the code, and remembering that lcΣB3 has to specifically remove comment text from within the configuration (on the page, not transcluded), I would say that lcΣB3 reads the raw page information, not processed output. Actually, thinking about it more, it has to read the unprocessed page, not anything processed, in order to have access to the config information. Thus, what the template outputs onto a rendered page is something that lcΣB3 never sees. — Makyen (talk) 04:57, 8 April 2014 (UTC)

You're right, an archiving bot has to look into the source code, otherwise how could various parameters be fetched? They're nowhere to be found after the templates are processed, or at least that's how it looks to me.

How reasonable would it be to simply request a protected page edit on User talk:MiszaBot/config, referencing this discussion at the same time? Thoughts? — Dsimic (talk | contribs) 05:09, 8 April 2014 (UTC)

After looking at the bot's source code, I can only confirm that it reads raw source code of talk pages to be archived. The configuration is fetched as code = mwp_parse(self.page.content), where mwp_parse() calls mwp.parser.Parser().parse(text, skip_style_tags=True), and mwp is a Python parser for wiki code. — Dsimic (talk | contribs) 05:30, 8 April 2014 (UTC)

(edit conflict) That sounds like the way to go. Given that you and Redrose64 did the work to find the fix, you should be the one to make the request. I was going to offer to do so, but that feels like taking credit for your work.

What you report from the source is what I recall seeing in my brief looks for other things. I also don't recall seeing any diffs which would indicate a change in that area.

As an aside: Even with that change, I would still use at least very similar wikitext. I would still want a forced TOC. We have to assume the possible need given that a section on a talk page might be very long. I would also still want the {{clear}} (after the forced TOC) in order to make the archive box not extend into the first section if the TOC is short (it is worse if there is no TOC). Given that Template talk:Linux currently only has three sections, you can play around with a preview of how it would look with and without the forced TOC and with and without the {{clear}}. I definitely prefer it with both. With the proposed change the }}__TOC__{{clear}} could be on its own line (until encountering one of the other misbehaving templates). — Makyen (talk) 05:39, 8 April 2014 (UTC)

To me, it's much more about making things better, than about credits. :)

I see what you're talking about, regarding the layout/formatting issues. How about making a "bundled" template for talk pages, which would consist of all things you've described? To me, it's always been strange that there's no template which provides both automated archiving configuration and a message which indicates that automated archiving is in effect on a talk page. You could create a template which would provide all of that – automated archiving configuration, automated indexing, a visible message, and various layout fixes. That way it could be easily deployed on many more talk pages, making the whole thing much better.

Thoughts? — Dsimic (talk | contribs) 05:54, 8 April 2014 (UTC)

Protected page edit added, User talk:Misza13 § Protected edit request on 8 April 2014. — Dsimic (talk | contribs) 06:22, 8 April 2014 (UTC)

(edit conflict) Request looks good.

There is such a template. It is the ClueBot III (CB3) config. CB3 also does archive indexing. I just find it quite limiting as to how it presents the archive box. I really don't like it when there are a large number of archives (see the top of this page, and imagine how that would look if it was not enclosed in a scrollbox). I also don't like a couple/few other things about CB3: A.how it handles placing sections when doing archives by date (see Help:Archiving a talk page#Automated archival for more info), B. it can take 4, or more, days to start archiving once the config is on a page)

I have planned on doing something quite a bit like what you have described for a while now. I first realized that there was a significant need for improvement to make it easier to add archiving a few months ago and started in with changes to Help:Archiving a talk page to make it such that people could just copy & paste a config. I have intended to both put complete config examples on that page and make a couple/few templates intended to be subst:'ed into a page. Having one actual template do it all would take coordination with the bots as the bots use the current templates to read their configs. Thus, without changes to the bots (unlikely), those actual templates must exist on the page. However, a template can be made that could just be subst:'ed in with a variety of possible config options.

One complicating factor is that a separate page has to be created if one wants to use HBC/legobot for archive indexing. However, the archive indexing by legobot (task 15) is basically dead, in that it has only run once in the last year (December 10, if I recall correctly).

However, at least for the archives by months subst:'ed template, I had intended to hold off until I, or someone else, had re-written the various archive box templates in lua. Some of them are running up against the maximum limit for "Expensive parser function count" (one expensive parser function per check for the existence of a page); at least one required changes just to keep operating into 2014. After I asked, and a bit of work by the people over at Wikipedia talk:Lua, it was found that lua can do the job without the Expensive parser functions. I have intended to use it as an excuse to learn lua. However, if you or someone else wants to do that work, go for it.

At least for a while, I have gotten tied down in re-working of WP:Cite4Wiki. Which ballooned way beyond by initial expectation/intention. But, I am getting to the point where I need a distraction from that, so I may whip up at least the complete configs for Help:Archiving a talk page#Automated archival and the subst:'ed templates, which probably are the biggest help for level of effort. Given that I already have complete configs which I have tested on multiple pages that should not take long. Ideas/effort are, of course, very welcome. — Makyen (talk) 06:35, 8 April 2014 (UTC)

I'd say you should go ahead and create such a "bundled" template, with whatever currently works, which would include configurations for bots and visual elements. That way no modifications would be required in archiving bots; down the road, you can always introduce new features, and even persuade authors of bots to implement required changes. It's all about introducing one improvement after another – what's a never-ending game, as we all know.

WP:Cite4Wiki looks like a very useful utility/plugin, and I know very well how easily (and quickly) things can go far beyond expectations; browser plugins are especially prone to such growth. Well, that's one more reason for a distraction in form of "bundled" template's initial version. :)

Just as a note, learning Lua isn't that high on my to-do list, :) but I'll think about it. — Dsimic (talk | contribs) 07:22, 8 April 2014 (UTC)

I have no control over the content of that template, as I am not an admin. I agree that an edit request would be the correct way to approach this, so it appears that it's being handled.

As for the "unified archive template", Legoktm and I once discussed plans to annex the archive indexer to lcsb3. As of now, we have come up with nothing worth enlisting on-wiki attention, but I will keep this issue in mind. →Σσς. (Sigma) 07:11, 8 April 2014 (UTC)

@Dsimic: Yeah, I was looking at the lua thing as just yet one more programming language. But, that is kind of what I told myself for the JavaScript needed for Cite4Wiki 8-). At least that has been useful for a few other things, and I expect will be useful in the future. As for Cite4Wiki, I find it very useful for generating citations. I'm going to be interested in some alpha testers in the (relatively near) future. If you are interested, watch the talk page there as I will ask for them there. Or tell me and I'll ping you. However, I still have a ways to go to get to that point.

I will get on the subst:'ed template(s) for putting basic archive configs on the page this week. You're right, it is something that would be helpful for a good number of people and would probably reduce the amount of support effort needed.

@Σ: I read what I believe was the initial contact of you approaching Legoktm about bundling indexing into lcsb3. Given that legobot task 15 is basically dead, I was/am hopeful that something would come of that. If it does happen, one thing that would be quite helpful is to change the logic wrt. the bot deciding that the page is acceptable to overwrite from only if the magic string is on the page to magic string or the page does not exist (and is in the default location). There are a large number of indexer bot configs that have never been acted upon by the bot because editors did not create the index page with the <!-- Legobot can blank this --> text. — Makyen (talk) 08:43, 8 April 2014 (UTC)

I hear you; back at the time, I've ditched the idea of learning JavaScript as it simply requires total devotion and a lot of time if you really want to know it – and nowadays time seems to be the only limiting factor for individuals willing to learn more. :)

I've added myself into the list of Cite4Wiki watchers, and I'll try to help as much as I can. It surely looks like a very useful utility; if I may ask, have you considered to implement it as a Greasemonkey script instead? That way you might be able to save yourself from the nightmare which producing Firefox plugins tends to be.

Having a "bundled" archiving template would be a very good thing, and it would be great if you'd have the time for making an initial version of such a template. — Dsimic (talk | contribs) 17:53, 8 April 2014 (UTC)

Please comment on Talk:Florida's 13th congressional district special election, 2014

Greetings! You have been randomly selected to receive an invitation to participate in the request for comment on Talk:Florida's 13th congressional district special election, 2014. Should you wish to respond to the invitation, your contribution to this discussion will be very much appreciated! If in doubt, please see suggestions for responding. If you do not wish to receive these types of notices, please remove your name from Wikipedia:Feedback request service. — Legobot (talk) 00:04, 9 April 2014 (UTC)

Threads deleted, not moved

Hi, with this edit, two threads were deleted but not saved in either of the two archives named in the edit summary. --Redrose64 (talk) 11:15, 11 April 2014 (UTC)

It appears to have just been 2 standard API errors in a row. Ideally it won't happen again. →Σσς. (Sigma) 19:13, 18 April 2014 (UTC)

need protection please

Hello sigmabot, could you add protection templates in those articles listet above, as you did in List of modern-day Muslim scholars of Islam? Thanks. Rinfoli {*Di§cu$$ with me"#} 18:22, 21 April 2014 (UTC)

This is not the bot's talkpage - although I quite understand why you might think it was. The bot is currently disabled, the bot owner is busy with other things, and the protection templates are not essential because they do not enable protection; they are merely a handy symbol indicating that protection is present. --Demiurge1000 (talk) 20:20, 23 April 2014 (UTC)

@Demiurge1000: Er, it is the bot's talk page. Check out what redirects here. --Redrose64 (talk) 16:20, 24 April 2014 (UTC)

Please comment on Talk:Australian Senate special election in Western Australia, 2014

Greetings! You have been randomly selected to receive an invitation to participate in the request for comment on Talk:Australian Senate special election in Western Australia, 2014. Should you wish to respond to the invitation, your contribution to this discussion will be very much appreciated! If in doubt, please see suggestions for responding. If you do not wish to receive these types of notices, please remove your name from Wikipedia:Feedback request service. — Legobot (talk) 00:06, 25 April 2014 (UTC)

Bot malfunctioning

See https://en.wikipedia.org/wiki/Wikipedia:Administrators%27_noticeboard/Incidents#.E2.80.8ELowercase_sigmabot_edit-warring_with_itself_on_Turkey for details. Tutelary (talk) 06:42, 20 April 2014 (UTC)

I've disabled lowercase sigmabot, as it was edit warring with itself at Turkey (edit | talk | history | protect | delete | links | watch | logs | views). Feel free to restart it once you've fixed the bug. Best — Mr. Stradivarius ^{♪ talk ♪} 11:09, 20 April 2014 (UTC)

Thank you. I will work on this when I have time. →Σσς. (Sigma) 00:25, 21 April 2014 (UTC)

@Mr. Stradivarius: if something like this happens on an odd page it's worth considering {{Nobots}} instead of disabling. All the best: Rich Farmbrough, 18:36, 25 April 2014 (UTC).

Older unsigned and undated messages

I noticed that there were several undated and unsigned messages (some from 2011) that were not be archived. I added the {{unsigned}} template to them. Just curious, will the bot archive them now or should it be done manually? Quana. CBWeather, Talk, Seal meat for supper? 10:49, 27 April 2014 (UTC)

(talk page stalker) The short answer is that, no they will not currently be archived. lowercase sigmabot III (lcΣB3) will not archive a section without a date/time formatted in the strict syntax it expects. In order for the date/time to be considered "properly formatted" it must include text indicating that the timezone is in UTC. Specifically it should include the text " (UTC)". Unfortunately, the {{unsigned}}, {{unsigned2}}, {{unsignedIP}}, and {{unsignedIP2}} templates do not automatically add the " (UTC)" text. As an example, the date/time stamp after your signature on your post above is "10:49, 27 April 2014 (UTC)". It would not be considered valid if it was "10:49, 27 April 2014".

To be more exact, and having (briefly) looked at the source code, the date/time stamp must match the regular expression:

"\d\d:\d\d, \d{1,2} (\w*?) \d\d\d\d \(UTC\)"

If it does not match that regular expression, then it will not be found as a date/time. — Makyen (talk) 11:23, 27 April 2014 (UTC)

OK. I'll archive them myself. Masi cho. CBWeather, Talk, Seal meat for supper? 18:56, 27 April 2014 (UTC)

Archive counter not expanded

Hi, please amend Lowercase sigmabot III so that arguments like %(counter)d are replaced with the true integer value. See the error message at the top of the recently-created Wikipedia talk:WikiProject Military history/Archive 125 and my fix. --Redrose64 (talk) 07:13, 27 April 2014 (UTC)

This is actually hard to implement given the current structure of the bot. I believe Makyen had come up with a solution at some point in my talk page archives, but I will look into this later. →Σσς. (Sigma) 23:50, 28 April 2014 (UTC)

Yep. That was lcΣB3: does not substitute variables in archiveheader field in February. This is an almost identical situation. I have applied the same workaround to Wikipedia talk:WikiProject Military history. — Makyen (talk) 00:29, 29 April 2014 (UTC)