Wishlist on organising the information with transclusion

I have been trying to rack my brains and come up with ways we can organise our information in more useful ways. What I would like to see would be a system whereby:

  1. Data by Assessment We could easily find all of the articles within a given quality category (A-Class or whatever), and count them. Ideally we would even be able to organise them into categories, so we could have something like this list, but generated automagically. That way if User:Joe Smith from WikiProject:Herbaceous borders (i.e., not one of us) went in and changed the assessment rating of the article Gertrude Jekyll from B to A, it would automatically appear on the list of A-Class articles (in the right section?). If this were done properly, Joe Smith could edit the WikiProject's worklist and then our listing would also change, along with the list of A-Class articles, and perhaps a listing in Core Topics or Vital Articles too.
  2. Data by article We should also be able to ask the question, "What is the assessment of the article on DNA Resequencer?" and get an answer quickly and easily. Ideally we should also be able to get other data such as WikiProject, date of assessment and comments (data we currently record) as well as importance of the topic and possibly category (data we may be recording in the future). One simple way of doing this would be to generate lists of articles in given fields, ideally as an article tree like this one (only much more slick!), and have links to the appropriate place in our assessment tables. Perhaps a better (but challenging) way for the long term would also involve automagic, we could have a set of templates where we give it the page name, and out come whatever metadata you need. In other words, you type {{Assessment|DNA Sequencer}} and it would deliver the variable "A-Class." This would be extremely useful for things like the article trees, where in place of a link to the article alone we could have a link to the article in a colour representing the assessment (Oh, it's red so it's a stub, sort of idea).

I am very limited in my technical skills with transclusion - I can only work on the principle that if I keep typing I will eventually get something that works. Well, almost, anyway - at least you can take a look. I have created a template for A-Class articles called {{User:Walkerma/Sandbox}}, and used it in this test table for listing the A-Class articles. It does require reordering the info to put the assessment in the first column. I think it's very easy to use, so Joe Smith can easily see how to edit the listing for Gertrude Jekyll, and it was easy to change the table from our current style. The only problem was that the comments column didn't come out, I don't know what I'm doing wrong there. Well, actually I just don't know what I'm doing, period. I have produced this just to show the sort of thing I am thinking of for item 1 (data by assessment) on my wishlist above.

Please give comments, and if you know how to produce the relevant automagic I will be spellbound! Or perhaps you can see other things we need to have, or better ways of achieving 1 & 2 from the above list. I would like to resolve this and build in the automation all now, before we start asking the projects for more lists, because once we have several thousand articles listed it will be very hard to change the system. Walkerma 05:09, 22 April 2006 (UTC)

It can be done, at least some of it. Let's follow how Tropical cyclones does this:
  • All articles from the WikiProject contain the {{hurricane}} template, which contains an optional parameter, class. If the class is defined, the template calls on another template to display the assessment (e.g. Hurricane Nora's assessment is B-Class).
  • The {{hurricane}} template contains a category that is automatically updated; a bot can be programmed to grab the category daily and make any changes if necessary. You may want to ask Oleg Alexandrov for help, as he programmed Mathbot to produce things like these.
  • Also, the inclusion of the {{B-Class}} template triggers an event that is recorded in Special:Whatlinkshere/Template:B-Class. As you can see, Hurricane Nora is listed down there with (inclusion) after the name. That can also help making lists of articles.
  • However, making assessments by article is incredibly difficult. In order to be able to grab that data quickly, we might need something along the lines of article validation enabled, as there is no way a template can use information in the database not passed as a parameter. Perhaps the folks at the Village Pump may have ideas, but I seriously cannot think how (perhaps the category scheme and a Toolserver account may even be needed). Titoxd(?!? - help us) 05:27, 22 April 2006 (UTC)

Thanks Tito, that's very helpful. I've tried understanding how WP:Cyclones do this and I end up going round in circles, but perhaps you could set up the templates for us to do that. For the data by article, one thought that occurred to me was that we could maybe insert a category onto the article page itself that says Category:WP1.0 Places {{WP1.0 Assessment}} or somthing like that, which would place it in a category such as Category:WP1.0 Places A-Class. A couple of things would be needed to make this workable, though:

  • Make it so that the category can be read from the assessment
  • Make the category invisible to someone reading the article, it would only show up while editing part, but it could be read by bots and the like. (Is that what noinclude does?)

Is that a workable method? Walkerma 05:41, 22 April 2006 (UTC)

Well, as you can see, that's similar to what WP Cylclones does, because we include assessment categories inside {{hurricane}}, and it produces subcategories of Category:Tropical cyclone articles by quality. Categories cannot be hidden, the only thing noinclude does is that it prevents categories from being included when a template is transcluded, to categorize the template page itself. Let me go grab a copy of the template and dissect it... Titoxd(?!? - help us) 05:45, 22 April 2006 (UTC)
Here it is:
{| class="messagebox standard-talk"
|-
|[[Image:Hurricane Andrew Landfall.jpg|center|50px|Hurricanes]]
| align="center" | ''This article is part of [[Wikipedia:WikiProject Tropical cyclones]], a project to systematically present information on [[tropical cyclone]]s and storm seasons. If you would like to participate, you can choose to edit the article attached to this page (see [[Wikipedia:Contributing FAQ]] for more information).''

Introductory information, links, pretty much what the ordinary WikiProject templates do.

{{#if:{{{class|}}}
  |
{{!}}-
{{!}} {{{{{class}}}-Class}}
{{!}} align="left" {{!}} ''This article has been rated as {{{class}}}-Class on the [[:Wikipedia:Version 1.0 Editorial Team/Assessment|assessment scale]].''
}}

This part receives the class parameter called from the page; it displays the appropriate class template on the screen, and marks the inclusion of the template in Special:Whatlinkshere.

<!--
-->{{#if:{{{assessed|}}}
  |<tr><td>[[Image:Exquisite-kfind.png|28px|center|Peer review]]</td><td align="left">This article has been [[Wikipedia:WikiProject Tropical cyclones/Assessment#{{PAGENAME}}|assessed]] by editors of the WikiProject.</td></tr>
}}<!--
-->{{#if:{{{merge|}}}
  |<tr><td>[[Image:Merge-arrows.gif|48px|center|Merge talk]]</td><td align="left">The possible merging of this article has been [[Wikipedia:WikiProject Tropical cyclones/Merging#{{PAGENAME}}|discussed]] by editors of the WikiProject.</td></tr>
}}<!--
-->{{#if:{{{TCCOTF|}}}
  |<tr><td>[[Image:Crystal 128 kuser.png|28px|center|Collaboration]]</td><td align="left">This article has been selected as the [[Wikipedia:WikiProject Tropical cyclones/Collaboration|Tropical Cyclone Collaboration of the Fortnight]].</td></tr>
}}<!--
-->{{#if:{{{TCCOTF-old|}}}
  |<tr><td>[[Image:Crystal 128 kuser.png|28px|center|Collaboration]]</td><td align="left">This article has been previously selected as the [[Wikipedia:WikiProject Tropical cyclones/Collaboration archive#{{PAGENAME}}|Tropical Cyclone Collaboration of the Fortnight]].</td></tr>
}}<!--
-->{{#if:{{{TCCOTF-nom|}}}
  |<tr><td>[[Image:Crystal 128 kuser.png|28px|center|Collaboration]]</td><td align="left">This article is a current nominee for the [[Wikipedia:WikiProject Tropical cyclones/Collaboration#{{PAGENAME}}|Tropical Cyclone Collaboration of the Fortnight]].</td></tr>
}}

This part adds merge, collaboration and assessment notices in new rows, which helps avoid using additional templates.

|}<noinclude>
[[Category:Hurricane templates|{{PAGENAME}}]]
[[Category:WikiProject banners|{{PAGENAME}}]]
[[Category:Templates using ParserFunctions|{{PAGENAME}}]]
</noinclude><includeonly>{{#switch:{{{class}}}
 |FA=[[Category:FA-Class hurricane articles|{{PAGENAME}}]]
 |A=[[Category:A-Class hurricane articles|{{PAGENAME}}]]
 |B=[[Category:B-Class hurricane articles|{{PAGENAME}}]]
 |Start=[[Category:Start-Class hurricane articles|{{PAGENAME}}]]
 |Stub=[[Category:Stub-Class hurricane articles|{{PAGENAME}}]]
 |Dab=[[Category:Disambig-Class hurricane articles|{{PAGENAME}}]]
 |Template=[[Category:Template-Class hurricane articles|{{PAGENAME}}]]
 |Cat=[[Category:Category-Class hurricane articles|{{PAGENAME}}]]
 |=[[Category:No-Class hurricane articles|{{PAGENAME}}]]
 |[[Category:Unassessed hurricane articles|{{PAGENAME}}]]
 }}</includeonly>

This part is the magic part: It receives the class parameter from above, and if the parameter matches a pre-set expression (e.g. class=A, it lists it on Category:A-Class hurricane articles. The noincludes are used to prevent Template:Hurricane from being categorized into the assessment categories (and to mark the Template page with particular categories), and the "includeonly"s to guarantee that the target pages are categorized by assessment only.

Does that help? Titoxd(?!? - help us) 05:57, 22 April 2006 (UTC)

Yes, that is helpful, I have the general idea now, and it sounds like something along these lines would be excellent for this project. Is there a way to generate a worklist-type list of the type we use here (or its equivalent? In other words, could we generate a table of all the Atlantic hurricanes that lists them by year and includes the assessment in the table along with the date of assessment and the comments? Thanks, Walkerma 06:14, 22 April 2006 (UTC)

Since we pretty much didn't like worklists (or no one proposed them), we just stuck to the category system. There's two ways I can think of doing so: through CategoryTree (but it needs some modifications to the tool to work), or by a specially-programmed bot that does so. By year, there could be some cross-category listing with Category:Tropical cyclones by season, but that might be two different tables altogether. Titoxd(?!? - help us) 06:27, 22 April 2006 (UTC)

I've been asked by Titoxd to comment here. I am not very sure I understand in detail what you guys are planning in here, there is a lot of complicated stuff to read in this page, but if you need a bot to crawl through a list of categories and generating some lists/reports based on some criteria, I could do that. If you want the bot to also edit a lot of pages, that is also doable, but would require user supervision obviously. Oleg Alexandrov (talk) 00:44, 23 April 2006 (UTC)

Mostly, yes, crawl through categories, and fill out a template parameter based on the results of the crawl. Titoxd(?!? - help us) 00:48, 23 April 2006 (UTC)
Got it. Again, appears doable, but if you want the bot to daily modify some template(s)/articles, somebody's got to watch over the changes everyday to see if the bot did not screw anything up (bugs happen, server problems sometimes, etc). User supervision would be the hardest part. Once it is decided what precisely you want to, I could work on it. Oleg Alexandrov (talk) 00:56, 23 April 2006 (UTC)
Sure. We'll discuss it some more here, and we'll contact you back, ok? Thanks! :) Titoxd(?!? - help us) 00:58, 23 April 2006 (UTC)
I don't understand the technical stuff fully, but if it means what I think it does, then it is excellent! Can you clarify/comment on some things for me?
  1. We need to make sure that whatever system we use is easy for the Joe Smith from a typical project (mentioned above) to go and add a new article or change an assessment in that table EASILY. Viewed from "Joe's" perspective, would this type of edit be easy?
  2. A result of #1 would be that if possible IMHO we need a bot (or something) that changes the category to match the worklist entry, rather than one that changes the worklist entry to match the category. If this isn't possible, we'll manage, we'll just have to make it clear in our posting to the projects that they edit the category, not the table.
  3. Will it mean us having to add a WP1.0 category for every article on our lists? We may get some flak for this, people will complain we are "spamming" Wikipedia with "pointless categories." There is a small minority of people who are strongly opposed to WP1.0 for some reason. If there is some way to add a "hide" command to the cats it would be nicer, but if not, we'll live with it.
Please can you comment on the above? However, even if we can't do any of the things I list there, the benefits of this automation are huge, and personally I'm extremely grateful to both of you (Tito & Oleg) for taking the time to look at this. I'm really glad we have people working on WP1.0 who understand the more sophisticated techniques. As far as I can tell, this will allow us to move from handling a couple of hundred articles with a month or two lag time, to handling several thousand articles with only a day or two's lag time. Exactly what we need! (If I'm misunderstanding things, please tell me!) Oleg, I think we can set up people to monitor the changes, at least every couple of days if not daily. Thank you so, so much! Walkerma 03:34, 23 April 2006 (UTC)

(\r) The change would be quite easy to do; just like replacing Start-Class for B-Class in an article's talk page.

As for making the bot read the tables instead of the category pages: Oleg was a bit concerned about checking for errors when the bot writes to a page. Making a bot read the categories and write to the article assessment tables makes it easier to spot errors by the bot via watchlists; unfortunately, there's no way to "watch" a category for additions, so it would be harder to check that way, as edits would be scattered among perhaps thousands of talk pages.

But yes, with this method it is necessary to have categories. However, the categories would be put on talk pages, like the rest of Wikipedia meta-data. Titoxd(?!? - help us) 03:50, 23 April 2006 (UTC)

Thanks, and I take your point about the problem of spotting errors. Talk pages are definitely less intrusive, I had forgotten that Cyclones uses the talk pages rather than main articles. I think "Joe" could easily do that. This all sounds like it will be very effective, thank you again! Walkerma 04:20, 23 April 2006 (UTC)
I was bold and picked a few WikiProject templates to use as testbeds for this expansion. Currently, WikiProject Tropical cyclones, WikiProject Chemistry, and WikiProject Meteorology will use the expanded templates, then if it scales, I'll begin adding the optional parameter for more projects. Titoxd(?!? - help us) 01:25, 24 April 2006 (UTC)
Thank you, I was hoping you might do that! (BTW, WPChemicals is much more active with assessing than WPChemistry, that may be a better choice.) As usual, I want to raise some issues we need to consider such a big change, because I want to get things right from the start.
  • Will Oleg be writing a bot to produce a worklist and possibly a tree as well? I'd love to see a bot-generated worklist for the cyclones project. Would the bot also be spitting out a list of assessment changes for us each day?
  • Can we have the bot record in the worklist the date of any assessment change? Within a project like Cyclones you may have people who say, "We haven't assessed 19th Century Typhoons recently, I'll do them again" but we can't do that at 1.0. If we have assessments that are 6 months out of date they may be way off, a date of assessment might help flag problems, particularly if all of the articles (say) on Ancient Greece haven't been done for a year.
  • Can the bot alert us more obviously (a listing in bold?) if an assessment goes down rather than up? I think generally articles improve over time, and although this downgrading will happen legitimately, it will sometimes flag assessment vandalism. And what if an article suddenly jumps from Start or Stub to A, isn't that suspicious too - could we catch that?
  • How are we dealing with categories? I think we should start by using the top level categories, I think based on this (please vote!) Option C looks like the winner, viz. Arts, Language & Communication, Philosophy & Religion, Everyday Life, Social Sciences & Society, Geography, History, Engineering (or Applied Sciences) & Technology, Mathematics, Natural Sciences. So rather than creating one big category of B-Class articles, we should have it listed as "Natural Sciences A-Class articles" and the like. Or were you planning on using the WikiProject name as (in effect) a category?
  • Can our worklists still contain columns for "importance" (to be added) and "comments"? These would not normally be edited by the bot but by people.

Sorry if this creates lots of work for Oleg! I'm really excited about the possibilities this automation will offer us. Thanks again, Walkerma 02:18, 24 April 2006 (UTC)

    • Replying one by one again:
      • Trees: Not sure; I haven't seen Mathbot do it, but it might if it is coded properly. As I'm not 100% sure of the specifications for the trees, could you ask him on his talk page and tell him how they work?
      • Cyclones worklist: We haven't had a need for it yet, but we have an Assessment subpage to discuss promotions to A-Class, and that has scaled fairly well. It sure would be interesting to see, although I don't know if we would use it.
      • Assessment dating: I believe Mathbot records the date an article is added to a category, but I'm not sure how the algorithm works exactly. It might crawl the category for new pages, so it might just record the date if it sees a change in the category, or it might crawl the category and dig up the date it was added from the history (doubtful).
      • Alerts: I know Mathbot does warn about category removals, and I think that a bit of code to detect more than two levels raised at once is doable.
      • Categories: Effectively, we can make a category tree based on Category:WikiProjects or something different; but I cannot see a way we can just lump all the articles onto Category:B-Class Wikipedia 1.0 articles and figure out from which WikiProject they came from (we can use it as a parent category, though). I used "A-Class chemistry articles" and "A-Class meteorology articles" to match up with the "A-Class hurricane articles" category we were using, and then add some sort of marker that the category belongs to a particular WikiProject.
      • Other columns: Sure, I just think we would need to design a template (easy) with named parameters, and just have Mathbot fill one of the variables (perhaps two, if you make the date a variable) and leave the other ones intact.
      • Right now, I'm just setting the infrastructure so we can contact Oleg to adjust his bot... so it takes a bit to get everything figured out, but it'll be up soon. :) Titoxd(?!? - help us) 21:17, 24 April 2006 (UTC)
Thank you once again! Regarding the date, the date the bot finds the change is perfectly fine, assessments change over months, not hours. I think worklist tables will be essential, as we have to see how the articles look in one particular area. We have around 570 articles in chemistry that we have designated within the project goals as being on significant topics. Around 555 of these have been assessed using the FA/A/GA/B/S/S system, so these should provide a good dataset for the bot to try. Since you have exams now/soon, and I have them looming, I suggest we don't introduce this new system until about a month from now (and my grades will just be in). I suggest we gear up to contact all the active projects (Toolserver list) at that time, including the ones we haven't done so far. In the meantime we can be adding assessment values into the articles and testing the bot. We may also want to get a basic system up for Release Value Qualifying, so we can also be asking the projects to nominate their best articles for inclusion in the WP0.5 release. I'll contact Oleg directly. Thanks, Walkerma 06:17, 26 April 2006 (UTC)

Before coding a new bot, you may want to see if the code I did already can be used with mods or as a basis... I walk the category dump to get the info. See User:Lar/ClassificationTableGen ++Lar: t/c 03:38, 27 April 2006 (UTC)

New category

In order to accomodate all of this, I've created Category:Wikipedia 1.0 assessments to simplify the gathering process. Subcategories of that page should correspond to the top-level categories of WP:1.0 (whenever they are set to be decided), and the subcategories currently there should be pushed down to be sub-sub-categories of that. One level deeper are the actual assessment categories, so they're categorized by area, WikiProject and quality assessment. The category is not completely populated; I'll try doing as soon as possible, as I'm in the middle of finals right now. Questions? Comments? Titoxd(?!? - help us) 06:49, 25 April 2006 (UTC)

Looks good to me! Walkerma 06:17, 26 April 2006 (UTC)

{{genre}}

Just to mention, if anyone's interested, that I've made some modifications to the basic assessment level code on WikiProject Music genres's {{genre}} which can be seen at talk:ambient music. --MilkMiruku 11:37, 26 April 2006 (UTC)

passing more info to save time?

might it be an idea if the templates (i.e. {{genre}}) were modified to pass on more info such as Importance, Date, Version, Comments (i.e., {{genre|class=B|importance=High|date=May 4th 2006|comments=Needs a layout cleanup, references, images}}), then Mathbot would be able to grab it when compiling the assessment lists? or not? nb. to combat abuse, the bot could check for certain keywords, dodgy dates, invalid importance levels, and, as previously mentioned, assessment levels that change to quickly. --MilkMiruku 00:13, 4 May 2006 (UTC)

I have seen the future for WVWP, and it looks like this!

Need: The article's importance, regardless of its quality

Top Subject is a must-have for a print encyclopaedia
High Subject contributes a depth of knowledge
Mid Subject fills in more minor details
Low Subject is peripheral knowledge, possibly trivial

I got very excited when I stumbled across this Math worklist being created for WP1.0. I was first pleased just that the project was putting together a list for WP1.0, then I get very excited when I saw that they also include a column in their table for Need, defined as "The article's importance, regardless of its quality". They have created a lovely set of templates in various shades of purple to indicate this, shown left. Are they psychic, or are they just reading the postings on this page about requesting lists on importance? This seems like a very simple system that anyone could understand. I would like to suggest that we use their system as a model, and use their templates too. We should

  • Incorporate "need" into our tables as a column, just like they do, and
  • Request all of the WikiProjects rank not only the quality but also the need.
  • Use the term "key article" for all of the "Top-Class" need articles.

Comments, anyone? (By the way, Tito, I will get onto checking our other project contacts tomorrow night, I just wanted to get all this stuff on the table, no pun intended). Walkerma 05:24, 22 April 2006 (UTC)

I agree, this has a lot of potential.
As food for thought, possibly the topics in the current core list would be in the "highest-need" class, and so forth. Maurreen 18:10, 23 April 2006 (UTC)
But some projects might not have any articles that are "high-need". Maurreen 19:08, 23 April 2006 (UTC)
Yes, you're right, certainly when the releases are small. Clearly a project like History would produce far more top need articles than the (inactive) History of Poland project. It would be the job of the projects to assign their relative importance within that subject area - then our job here to assign the importance as a whole. So we might pick the top ten from the Poland group, but the top 500 from the history project. Any thoughts on how best to do that - to reconcile the ranking of "high" from a project with our judgement of "low" importance for the subject as a whole, without upsetting the low-ranked WikiProjects? Walkerma 01:35, 24 April 2006 (UTC)
Hm, that looks awfully familiar. But no, seriously, I'm glad our setup is propagating across Projects. Note that the CVG list rescaled yesterday, adding a new "disputed" category and bumping almost everything down one; it cut the number of top-listed articles from nearly 50 to 7. Nifboy 19:54, 24 April 2006 (UTC)
Indeed, I pinched it from CVG when creating the page. --Salix alba (talk) 17:16, 15 June 2006 (UTC)
Ah, I knew I had seen it before, but I couldn't remember where. I'm not sure we would need a +/- scale, as you have there, but I can see the use for a disputed tag... or if it is disputed, just leave blank? Titoxd(?!? - help us) 21:02, 24 April 2006 (UTC)
Ditto about the letter scale from A - F. I do like what the Math WP's done. —Mirlen 21:22, 24 April 2006 (UTC)
The +/- scale is there because B-class is a lot of articles; pop-culture topics don't naturally acquire the references necessary for A-class. The A-F (A-D up until yesterday) scale is arbitrary and, I think, close enough. I'm not sure why F was added to the scale yesterday. Nifboy 21:32, 24 April 2006 (UTC)

There seems to be some conflicting advice on this. Wikipedia:Version_1.0_Editorial_Team/Release_Version_Criteria#Importance_of_topic says "Importance must be regarded as a relative term. If importance values are applied within this project, these only reflect the perceived importance to this project. An article judged to be "Top-Class" in one context may be only "Mid-Class" in another." Above we seem to be saying that we should take a wider view of importance. So, do I grade The KLF as "Top" (it's the most important article in our WP:KLF micro-WikiProject) or maybe Mid (I'll push for High! ;)) in the wider scale of things? --kingboyk 14:54, 27 May 2006 (UTC)

Excellent example! That's exactly because it's a relative term, i.e., it depends on the context. Within the KLF WikiProject, the eponymous article is obviouly Top-Class. Within the electronic music project it would probably be the same. Within a project on pop music as a whole, the KLF might only be considered "High-Class", while for WP:1.0 as a whole it would probably only be "Mid-Class". I would like projects to maintain their own priorities appropriate for that project. If there is a separate article about the jackets worn by the lead singer, you may consider that Low-Class even within the project, for example. When we judge importance for WP:1.0 based on these lists, we will in effect do a calculation of
(Importance of subject area) X (Importance of article within that subject area)
to come up with our own importance assessment. Walkerma 15:47, 27 May 2006 (UTC)
Thank you. I agree with your ratings across the different contexts, too. So, I'll grade within the context of our Project(s) and let you guys do the calculations! This is good to know for the rather-more-important Beatles WikiProject too so thanks again. (I'm using the KLF Project as a kind of testing ground for these things :)). --kingboyk 16:04, 27 May 2006 (UTC)

Wikipedia:WikiProject Music genres

Just to mention I've made some steps to getting people interested in article assessment for 1.0 on WikiProject Music genres, such as creating Wikipedia:WikiProject Music genres/Assessment and popping a note in {{genre}} (which appears on the talk page of most genres so shuold hopfully attract a few interested parties). Any other thoughts or tips on things we should be doing or trying out? --MilkMiruku 21:20, 24 April 2006 (UTC)

It looks nice. We have been talking above about adding the asssessment directly on talk pages via templates; do you mind if we add your WikiProject to the trial group? Titoxd(?!? - help us) 21:27, 24 April 2006 (UTC)
sure, that would be cool. one thing i was wondering about was how to keep it updated at Wikipedia:Version 1.0 Editorial Team/WPArts#Wikipedia:WikiProject Music genres, although it also worried me that listing the whole thing there might have potential to bloat it's length, given the number of music genre articles on wikipedia. --MilkMiruku 21:42, 24 April 2006 (UTC)
Well, right now, it has to be done by hand, which will hopefully change soon. Don't worry about the size; the more articles we identify, the better. Titoxd(?!? - help us) 22:17, 24 April 2006 (UTC)

Food for thought

Running a query of the WP:WVWP tables page, the project has assessed 733 articles to date, and I thought that was interesting, as there's several projects that have not replied to our contact. Have all of the Arts, Humanities and Hobbies projects been contacted? Titoxd(?!? - help us) 04:54, 25 April 2006 (UTC)

Offhand, I know there were a few very specialized CVG project descendents that were either recently revived or created (and not asked): Digimon, Nintendo, The Elder Scrolls, and the less active Nintendo Wars, Tycoon games, and Mortal Kombat. Nifboy 05:45, 25 April 2006 (UTC)
Well, if they're really recent, chances are that they are not on the Toolserver report of WikiProjects (and we still need to go through that list again). I'm contacting them now. Titoxd(?!? - help us) 05:51, 25 April 2006 (UTC)
Roughly speaking, we have contacted all of the projects on our original list. However being Wikipedia, this list is already hopelessly out of date, we are missing probably 100 or more projects. Once we update our list to the toolserver list there will be many active projects we will be contacting for the first time. Walkerma 12:38, 25 April 2006 (UTC)

Changes to assessment scale

I've done one change to the assessment scale, and to keep discussion at one place, see the main project's talk. Titoxd(?!? - help us) 03:17, 26 April 2006 (UTC)

Wikipedia:WikiProject The Beatles/Article Classification

You may not know that there is now a Beatles WikiProject. We have begun the long task of classifying our articles. I'm sure you'll want to have at least The Beatles, the four individual members, and the key albums in Version 1.0 so please add us to the list and get in touch for some cross-project co-operation. --kingboyk 03:08, 27 April 2006 (UTC)

Some thoughts

Having stumbled on these discussions, I'd like to offer a few comments on some of the proposed changes. Many of these are related to the experience of one WikiProject in particular (the curious reader can easily find out which one) and may just wind up being special cases that we'll need to work around.

Article rating via project banners: a clever idea which might work most of the time, but will come up against a number of issues:

  • Project size: a project with 500 talk pages to watch for changes might be able to do it; a project with 5000 will have trouble keeping track of changed ratings even with bot assistance.
  • Contentious articles: the average article on an inorganic compound is unlikely to be the victim of POV edit warring; articles like Iraq War, on the other hand, will probably see hourly changes in ratings. Certain projects have a disproportionate number of controversial articles to deal with.
  • Meaningfulness of ratings: with hard-coded worklists, there was some assurance that the ratings had been given by a member of the project; with ratings directly on the talk page, this would seem to no longer be true. Is talking about selection of articles by WikiProjects even meaningful if the ratings are given by random passerby?

Article importance ratings: a good (if likely to be contentious on a case-by-case basis) idea. However, I would think that assigning "top-level" WikiProjects greater responsibility here would be a mistake, as many (like the History one) are either totally defunct or mere groups of a few inexperienced editors. Priority should be, in my (obviously biased) opinion, given to working with larger and more active projects, even if they are further down in the hierarchy.

Comments? ;-) Kirill Lokshin 04:23, 27 April 2006 (UTC)

These are good thoughts. I agree that tracking ratings via talk pages isn't perfect, but it's at least a workable system. Personally the thought that our little handful of active people at WP1.0 might have to track the assessments of 5000 articles by hand is scary! For the moment, we'll have to put up with flaws, and probably check the most contentious articles by hand at the time of release (as we'd probably do anyway). One saving grace of User:Mathbot/WP1.0 is that it (he? she?) will be generating a daily log of changes to assessments. I have requested that it put in bold any changes that look suspicious (up >1, or down any levels) so we can check those carefully. Another help is that a random passer-by would have to look at the talk page and have some knowledge to work out how to make such edits (most vandals fail on both of these). It may be that projects will want to keep their hard-coded worklists anyway (I suspect at WP:Chem we probably will, it's so much a focal point of the project).
As for importance, I meant that even a "high-class" or mid-class importance article as judged from the whole of history might be something like Theodosius I, this is likely to be as important as Bolesław I the Brave which may be rated as "top class" by a History of Poland WikiProject. We will need to adjust importance to allow for this. My experience from contacting 100 projects or so is that they will probably use such ratings differently - some may come up with 100 top-class, others none. At WP1.0 our job will be to interpret this and say "We have space for about 10 articles on the history of Poland. Let's pick the 5 they list as top-class and the six A-Class articles ranked as high-class." I take your point about active vs. inactive projects, but we will have to try and judge relative importance between subjects. We won't please everyone, I'm sure, but we can at least pick the most important within each subject area. Thanks for your comments, Walkerma 04:53, 27 April 2006 (UTC)
Thanks for the response. A few further comments:
  • By passerby I meant not necessarily common vandals, but the creators of new articles, people engaging in edit wars over them, and so forth. To the extent that the ratings are now distributed, this becomes less of a "Work via Wikiprojects" approach and more of a general validation scheme that happens to be tied to WikiProject banners.
  • The MathBot thing looks nice, but will it be able to generate per-project changelists? Obviously the major point of this approach to 1.0 is that the WikiProjects can keep track of rating changes, rather than forcing you to do all of it yourself; but this would require being able to find changes of interest easily.
  • On a side note, how will it deal with a talk page that has banners from two projects with different ratings?
  • Obviously history as a whole would have different "most important" articles than, say, history of Poland; and these would, presumably, be given priority. I'm drawing a distinction, however, between the abstract discipline and the related WikiProject. While it's important that someone pick the key history articles, for example, I would argue that the history project is unqualified to provide any significant ratings, by virtue of being defunct. (It's perhaps a too-subtle point more related to WikiPolitics than the actual articles in question; but we've never really worked out what the relationship between "hierarchical" WikiProjects is, and how we deal with cases where child projects are more active than their parents.) Kirill Lokshin 05:14, 27 April 2006 (UTC)
We are in fact planning on doing this process by individual WikiProject or at least by group of related WikiProjects (Chemistry only uses one common banner for all projects). I think we should see if Mathbot can give logs of individual project areas to the individual WikiProjects, I suspect that is easy to do once the log pages are set up. It would be up to the projects to monitor the details, but this might be a nice "service" we could offer them in return for their work. If two banners appear, I think it might appear on two lists, but that won't matter too much I don't think, though it may be amusing if assessments differ! I take your point about WikiPolitics; I've noticed the same thing in articles, often a far more specific topic gets a much stronger article than a general once (take arithmetic as an example from your field!). Many of our FAs are on topics IMHO too obscure or unimportant to be suitable for inclusion in a small WP1.0 release of (say) 3000 articles, yet many articles we want are at the start-class level. Thanks again, Walkerma 05:28, 27 April 2006 (UTC)
Very well, then. If you could perhaps let us at WP:MILHIST know once the MathBot logging is up and running, we'd likely be quite interested in converting to the new system. Kirill Lokshin 02:51, 28 April 2006 (UTC)

Status report

For those who just checked in to find all this discussion and activity overwhelming them, a quick summary is probably a good idea.

  • We are creating a new style of article table such as this one.
  • We are also organising assessments by organising articles into categories such as Category:A-Class chemistry articles, with the category appearing only on the article's talk page.
  • We are in the midst of testing out User:Mathbot/WP1.0, a bot which will crawl through all of these WP1.0 assessment categories once a day and look for any changes. The bot will then:
    • Update the article assessment table
    • Note the changes in a log - probably both a "global" log for us, and in a subject-specific log for the relevant WikiProjects (who will organise the assessments).
  • Once we have this system working, we'll be able to start contacting the WikiProjects, knowing that we have the necessary infrastructure to handle this.
  • Meanwhile back in Gotham City there are plans afoot to set up Nomination and Qualifying pages for V0.5 and V1.0 very soon. This means that our new contact with the WikiProjects can include a direct appeal to nominate their articles for inclusion on a CD or DVD.

Phew! Walkerma 01:38, 30 April 2006 (UTC)

We at the Beatles WikiProject have started our own classification effort (see thread above) using a similar table to yours but not - I think - identical. Could I ask for two things then please: 1) any help classifying our articles would be most welcome. 2) please don't develop a bot which can't read our table, or at the very least let us know asap what changes we need to make as we should hopefully soon be fleshing out our table to cover all our articles. I'd hate to see all that work go to waste. For now, then, I'll assume you will contact us when you're ready. --kingboyk 08:09, 6 May 2006 (UTC)
I want to try to bring what we did on WP:Beatles into the fold before it goes completely stale. We use more types of classification (as you can see at Wikipedia:WikiProject_The_Beatles/Article_Classification#Article_progress_grading_scheme Would anyone be upset if we added additional templates similar to Template:FA-Class to take on our merge/delete structure? If I do that, I think we can switch our current classification over to match the table structure used in Chem etc. (also start using the template it uses) so that they'd be unified. My previoius idea of contributing code to Mathbot wasn't a good one but before it all goes completely stale we better get on board. Thoughts? I am not sure if this is the one best place to ask though. ++Lar: t/c 23:08, 21 May 2006 (UTC)
My suggestion would be to have templates/categories for the non-standard classes, but to avoid trying to get Mathbot to parse them. See, for example, {{NA-Class}} and the associated Category:Non-article military history pages, neither of which is recognized by the bot; you could create the same thing (e.g. Category:The Beatles articles for merging) either from a {{Merge-Class}} tag, or, alternately, from a separate merge=yes parameter in the project banner template. Kirill Lokshin 00:30, 22 May 2006 (UTC)
I think Kirill is right. It's very easy to get the bot going, all you have to do is to create a Category:Beatles articles by quality similar to Category:Chemistry articles by quality, as well as the sub-categories like Category:B-Class chemistry articles (make sure you include the parent cat in the sub-cats). Then create a Beatles template equivalent to Template:Chemistry. Then you add this template and the appropriate assessment to the talk page of each article, something like {{Beatles|class=B}}. This should (I think) automatically create the list and log. You should probably check in at User_talk:Mathbot/WP1.0, and read User_talk:Mathbot/WP1.0#So.2C_how_do_we_use_it.3F, then mention what you are doing so Oleg is aware. Good luck! Walkerma 04:39, 22 May 2006 (UTC)

WikiProject Tycoon Computer Games

Hello. This is regarding the Wikipedia:Version 1.0 Editorial Team cooperation message you have recently placed on our wikiproject talk page. I am Iuio, member and representative for this wikiproject. We would like to know your criteria either for featured, A-class, B-class, or Good articles so we can make the appropriate recommandations. Please post your answer on our talk page. Thank you. (Iuio 01:23, 7 May 2006 (UTC))