Template talk:Authority control/Archive 12

Jump to navigation Jump to search
Archive 5 Archive 10 Archive 11 Archive 12


There are heaps of errors in articles: Lua error in Module:Authority_control at line 1192: attempt to concatenate field '?' (a function value). Example Irena Swanson. Recent edits to the module need to be reviewed. Johnuniq (talk) 10:54, 8 June 2021 (UTC)

Seems to be all articles in Category:Pages using authority control with parameters. Fram (talk) 11:25, 8 June 2021 (UTC)
I have reverted the change for now — Martin (MSGJ · talk) 12:15, 8 June 2021 (UTC)
Indeed, it was all pages using parameters and a Wikidata item (either implicitly by being in mainspace or explicitly with a |QID= parameter). This wasn't caught by the testcases since there weren't any tests that tested that specific combination. I've now added some tests (which were showing the same Lua error then but now aren't after the fix), and fixed the underlying bug in the sandbox (a straightforward one-character typo), so this should be ready to go live again (but I'd prefer a different template editor or admin, such as MSGJ, actually re-apply the change so I get a second set of eyes) * Pppery * it has begun... 13:13, 8 June 2021 (UTC)
I can't spot any problems in the test cases (but it is not exactly easy to cross-check the links because the format is so different!) If no one else can see any issues I will deploy shortly — Martin (MSGJ · talk) 14:12, 9 June 2021 (UTC)


I've just come across this template {{Authority control (arts)}} when @Fram: added it, replacing the previous {{Authority control}}, to Saara Hopea. The effect seems to be that the reader no longer sees the ISNI or SELIBR data, while the other three fields are unchanged. While I'm not sure who, under what circumstances, would use the ISNI and SELIBR, it seems odd and arbitrary to include them for most people but not for visual artists - especially as there are plenty of people who are visual artists but also write or do other activities for which non-visual-artist identifiers might be relevant. (And a visual artist could well write a book a couple of years after her article had been given this visual-arts-specific template had been added...)

Could someone point me to an explanation of why this separate template is needed, and how it benefits the reader? Is it part of a move towards a whole series of different specialised Authority control templates? I'm genuinely puzzled. Thanks. PamD 17:09, 10 June 2021 (UTC)

Amended above when I realised that although the templates have separate pages, the talk page is combined - I hadn't noticed I was being redirected. PamD 17:28, 10 June 2021 (UTC)
Link to relevant TFD. – Jonesey95 (talk) 19:16, 10 June 2021 (UTC)
@Jonesey95: Thanks. A long read, shedding quite a lot of light. I'm not convinced by the argument for the existence of this variant template - it looks to me as if the solution to the problem of "too many ids are displayed in {{Authority control}}" would be to have it collapsed by default, so that only a single line intrudes visually in the article. But the discussion, this time round at least, has been and gone.
Could I ask that something be added to the template documentation page at {{Authority control (arts)}} to explain this history and give a link to that discussion? If the talk page for the template existed, it would have a notice announcing that previous TfD, but the talk page redirects here instead. Should that be the case? I'm sure I won't be the only editor not involved in previous discussion about this who will be puzzled on seeing something on their watchlist being amended to change the template, and wanting to find out more about it.
And it seems surprising that the Art UK identifier doesn't seem to be included, for UK artists - is it considered too lowbrow a database to be of interest, or something like that, or because the unique identifier is based on the actual name? It might not be necessary for JMW Turner but would be useful for a John Smith. It's in Wikidata.
The template name is also misleading: "arts" usually includes music, literature, dance etc, but the template description specifies the much narrower "visual arts". PamD 22:50, 10 June 2021 (UTC)
I don't know why the template wasn't just merged here, with |1=arts or |arts=yes implemented as an option. That would have been much more sustainable. – Jonesey95 (talk) 00:00, 11 June 2021 (UTC)
A few points: the only IDs that can be included in Template:Authority Control (Arts) are IDs which are already included in the main AC template, as the Arts version is a wrapper, wholly dependent on the main template. If Art UK, British Museum, ... IDs are added to the main template, they will automatically also appear on the Arts template. Collapsing wasn't an option at the time the Arts version was created: if and when that happens, it might be useful to re-evaluate the need for separate subtemplates. But even then, it still would mean that on uncollapsing, one would get plenty of IDs which are of very little added value for specific articles (while they may be useful for other ones), and removing clutter while keeping the best bits is usually supported. Finally, it wasn't merged here because this would make the main template even more complicated (certainly if more such subtemplates would be created, e.g. one per country or language or so, or one for sports, one for music, ...), and would make maintaining and finetunnig the subtemplate impossible for anyone but the few people with the TE right here. Fram (talk) 07:27, 11 June 2021 (UTC)

DNB or GND or German National Library

I missed the discussions to change the display of authority control. Looking now, I couldn't find the information for the German National Library which formerly appeared with label DNB. I saw France etc. but no Germany. Nikkimaria told me it's hidden under "Integrated Authority File". How is anybody supposed to know that? Do you expect users to look at the template documentation? Would it be hard to - perhaps also - provide the link under "Germany" where it would be self-explanatory? --Gerda Arendt (talk) 15:27, 21 June 2021 (UTC)

I have no objection to changing the label and/or moving it to a different section, and doing both of those would be fairly straightforward. Listing it twice seems confusing to me, and would also be a little bit harder to implement. * Pppery * it has begun... 15:36, 21 June 2021 (UTC)
A simple solution could be: "Integrated Authority File (Germany)" or "Integrated Authority File (DNB)". Grimes2 (talk) 15:58, 21 June 2021 (UTC)
Added (Germany) to the sandbox. I decided that was better since one of the goals of the redesign was to avoid cryptic acronyms. If someone else prefers a different solution, that's fine with me as well. * Pppery * it has begun... 21:29, 21 June 2021 (UTC)
And deployed (noting for context that Gerda thanked me for my previous comment, which I took to mean that she agreed with the (Germany) addition. * Pppery * it has begun... 20:41, 22 June 2021 (UTC)

Several potential classical music properties

From the transclusions of {{Authority control Q}} ({{Authority control (custom)}} would be more accurate & appropriate, but anyway), there are several property IDs that don't exist in WD yet. I've gathered all the unique IDs below:

Of these, RISM ID (P5504) & FAST ID (P2163) already exist in WD, but are not in the module, and should probably be added here if no objections.

The rest I'm leaving here for anyone that might want to start the associated d:Wikidata:Property proposals.   ~ Tom.Reding (talkdgaf)  11:39, 21 May 2021 (UTC)

 RISM & FAST done   ~ Tom.Reding (talkdgaf)  14:06, 5 June 2021 (UTC)
I've added these to the redesign code (which is now back in the sandbox), giving RISM the label "RISM (France)" wikilinked to Répertoire International des Sources Musicales, and FAST "Faceted Application of Subject Terminology" with no wikilink, and placing both in "Other". @Fram: Do you have any better suggestions for labels and sections for these? * Pppery * it has begun... 16:43, 5 June 2021 (UTC)
FAST belongs in "other" indeed (if we need to have it at all, seems to bring little added value). RISM should perhaps go to a section "Music", although we could wait to create this until we have one or two additional music entries (We have Musicbrainz, but although this are a lot of links, most music articles have only one MB ID). Fram (talk) 07:38, 7 June 2021 (UTC)
@Tom.Reding: There is an ISWC Wikidata property, ISWC (P1827), but it isn't an identifier because the ISWC database isn't in a format that allows for linking to IDs directly. Perhaps a third-party site like ASCAP or BMI could be used to form links, but those sites might not have all ISWCs in their databases. Jc86035 (talk) 08:33, 27 June 2021 (UTC)
@Jc86035: thanks! Not sure how I missed that... What is wrong with these directly-linked IDs? Using the 2 examples @ ISWC (P1827), I'm able to link to both T-905.029.737-5 & T-900.214.198-2 using the third-party formatter URL (P3303). We use 3rd party formatters for other properties, and the main drawback I've seen is they might need updating every few years, and not their completeness, at least not for those URLs chosen as the main third-party formatter URL (P3303). I.e. if completeness is a problem, then it should be demoted/removed.   ~ Tom.Reding (talkdgaf)  12:23, 27 June 2021 (UTC)
@Tom.Reding: My understanding is that BMI and ASCAP have the same set of ISWCs, since they share data with each other's sites, but they might not have the full set of ISWCs because other organizations can also assign them. So if a work has 0% BMI control and 0% ASCAP control it won't be available on their sites. In practice the vast majority of contemporary works with enwiki articles will have an ISWC in BMI/ASCAP, though. I don't know if it would be appropriate to link to them here, even if it would be better than nothing. Jc86035 (talk) 12:29, 27 June 2021 (UTC)
@Jc86035: good to know for BMI/ASCAP. The current third-party formatter URL (P3303) for ISWC (P1827) is musicbrainz.org, though.   ~ Tom.Reding (talkdgaf)  12:54, 27 June 2021 (UTC)
@Tom.Reding: Depending on the context MusicBrainz could be worse since all ISWCs it has are added manually, so its dataset is almost certainly much smaller than the BMI/ASCAP dataset. I don't really know how P3303 is used in practice so I'm not sure how relevant the current values of it are to this template. Jc86035 (talk) 13:01, 27 June 2021 (UTC)

External links to library resources: RfC

A Request for Comment on external links to library resources, which relates to this template, has started: Wikipedia talk:External links#RfC: External links to library resources. Opinions, knowledge, and suggestions are sought. Please join in. SilkTork (talk) 10:29, 3 July 2021 (UTC)

Invalid containsVIAFID link

At Warberg IC, the "Authority control" section has "WorldCat (via VIAF)" link http://www.worldcat.org/identities/containsVIAFID/316392146, which is not valid. It results in "404: Document not found". While the Warberg IC wikidata:Q3486392 VIAF ID entry 316392146 links to valid http://viaf.org/viaf/316392146/. Shouldn't the links be the same? (both to viaf.org) --Prikryl (talk) 06:55, 4 July 2021 (UTC)

No |nocat parameter or similair

Hi, I was just looking at this templates documentation and there doesn't appear to be a nocat or equivalent parameter (specifically I was looking at Wikipedia talk:External links#RfC: External links to library resources, which currently is added in 89 different categories by my quick count)? If there is could someone point me to it, if not could a technical editor comment on the feasibility of adding one? -- Asartea Talk | Contribs 10:46, 8 July 2021 (UTC)

Why? Nikkimaria (talk) 12:05, 8 July 2021 (UTC)

Microsoft Academic

Microsoft Academic is (unfortunately!) shutting down at the end of this year. Do we have a suitable replacement we can add to this template? Perhaps Scopus? ProcrastinatingReader (talk) 09:47, 13 July 2021 (UTC)

Is there a known issue related to Authority Control template, and the Navboxes Top / Bottom format?

Evening all. Per an issue raised on Lionel Messi, and when checking this also occurs on Cristiano Ronaldo, it seems that when these players have Navboxes that nest inside others the Authority Control wont render and just gives the Template link. The only solution appears to be to place Authority Control above the Navbox dialogue? Koncorde (talk) 18:42, 23 July 2021 (UTC)

The Ronaldo page is in Category:Pages where post-expand include size is exceeded, which means that not all templates in the page will be rendered properly. The size of the page needs to be reduced. – Jonesey95 (talk) 19:23, 23 July 2021 (UTC)
That isn't the issue. I can move Authority Control to before the existing Navboxes, and I can add 3 such templates, and all continue rendering correctly. On the 4th it begins to create issues. But if I put Authority Control AFTER the Navboxes it wont work at all. Koncorde (talk) 21:59, 23 July 2021 (UTC)
I'm pretty sure that it is the issue. The limit report in the article's source currently says "Post‐expand include size: 2097107/2097152 bytes". It's 45 bytes under the limit, which means that a tiny change will put it over. It could be that the reordering of the templates changes the rendered page just enough that it gets under the limit. The solution is still the same. – Jonesey95 (talk) 22:42, 23 July 2021 (UTC)
I will try trimming some content down as testing removing a section entirely did enable the template at the end. Cheers Koncorde (talk) 00:08, 24 July 2021 (UTC)
Looking at the Special:ExpandTemplates for {{Authority control}} on Lionel Messi, there are some minor improvements that can be made, that together may shave ~500 B off the post-expand size.
  1. The {{EditAtWikidata}} module seems to be adding a large & possibly repetitive preamble (search for "edit" in the expand). The whole 1st line is ~1200 B long, but I don't know how much of it is un/necessary; like, do we really need the URLs in

    1.1 <div id="Authority_control_frameless_&#124;text-top_&#124;10px_&#124;alt=Edit_this_at_Wikidata_&#124;link=http&#58;//www.wikidata.org/wiki/Q615#identifiers&#124;Edit_this_at_Wikidata",
    1.2 aria-labelledby="Authority_control_frameless_&#124;text-top_&#124;10px_&#124;alt=Edit_this_at_Wikidata_&#124;link=http&#58;//www.wikidata.org/wiki/Q615#identifiers&#124;Edit_this_at_Wikidata"?

  2. All "Wikipedia articles with ..." categories have always bothered me since the wording seems redundant. Surely they can all be shortened to "Articles with ..."? (10 B x 23 cats = 230 B)
  3. There is a useless comment produced by Template:Authority control that can probably be omitted from the expand by moving around the <includeonly></includeonly> tags. (~45 B)
I haven't looked at any other template, but there is probably at least a little bit of fat that can be trimmed elsewhere.   ~ Tom.Reding (talkdgaf)  01:26, 24 July 2021 (UTC)

Regex of ADB and MA identifiers need a fix

Two valid cases detected as faulty identifiers:

  • The number at the end of ADB identifiers can now exceed 30000, e.g. Lawrence Bragg.
  • It is possible for MA identifier to have 4 digits only, e.g. Sangji University.

I have fixed Wikidata but could not change the module here. ネイ (talk) 14:59, 20 July 2021 (UTC)

Fixed. – Jonesey95 (talk) 16:18, 20 July 2021 (UTC)
Thank you! Looks good now. ネイ (talk) 14:25, 24 July 2021 (UTC)

Displaying the IDs & making it easy to copy and paste

Sometimes I copy IDs from Wikipedia in lieu of searching on various sites, so being able to see the number and just copy it is useful. I toggle between Wikipedia and Wikidata as sources for finding multiple IDs. The additional supporting information in Wikipedia articles often helps with confirming that it is in fact a match. BooleanColors (talk) 02:28, 27 July 2021 (UTC)BooleanColors

When you follow the link, you can easily copy the ID from the target website (or alternatively, you can find the ID at Wikidata, which often has two or three times as much IDs as Wikipedia anyway, while Wikipedia only has IDs from Wikidata and no additional ones). Displaying the IDs here as well has been removed after an RfC where most participants found it unnecessary or distracting and preferred the new format or something like it. To find out if the person at Wikipedia and the person at the ID target website are a match, you need to follow the link to that site anyway. Fram (talk) 08:17, 27 July 2021 (UTC)

Collapsing the template?

Am bumping into these large authority control templates on many pages, and the concept of navbox stacking of collapsed templates falls on the wayside. Can these things be default collapsed all at once? Thanks. Randy Kryn (talk) 18:18, 7 June 2021 (UTC)

There was no consensus to auto-collapse them (thankfully), nor to keep using the smaller version (sadly), so: nope. Mike Peel (talk) 18:24, 7 June 2021 (UTC)
Maybe after dealing with them the auto-collapse will become the popular option. Too many pages hide three or four templates in navbox cages, and then these things are now sticking out the bottom taking even more attention from well-made templates. I don't understand your 'thankfully', eyesores are eyesores and never Mark Twain shall meet (check out the bottom of Mark Twain's page, for example). Randy Kryn (talk) 18:29, 7 June 2021 (UTC)
I really hope not, auto-collapsing it would make it even worse than it is now. But given the track record here, that probably means it will happen soon. :-( Mike Peel (talk) 18:33, 7 June 2021 (UTC)
Readers who actually use Authority control links (any research on people using them?) would very likely have the smarts to click 'view' to expand them. Randy Kryn (talk) 18:42, 7 June 2021 (UTC)
Maybe existing users who wonder where it's gone on specific articles. But not all articles have it, so it would make it harder to see which articles have it and which don't. And it significantly reduces the chances of it being found by other readers who don't currently use it. Mike Peel (talk) 18:47, 7 June 2021 (UTC)
I agree with the original poster here. Please revert this horribly obtrusive and non-consensus-based expansion or face the likelihood that people will start removing these templates en masse from articles. Authority control should not be such a prominent part of articles. Having it expandable is ok (and was agreed to in the recent RFC). Having it default to the expanded state is not. Most readers have no use for this information. It just makes more distractions from the part they are actually here for, the text of the article. —David Eppstein (talk) 23:15, 7 June 2021 (UTC)
Came here because I also think this looks ridiculous. Authority control means nothing to ~99.9% of people reading an article. AllegedlyHuman (talk) 03:26, 8 June 2021 (UTC)

Anyone is free to go to the Village Pump and start a new RfC to get it collapsed, removed altogether, reverted to the old version, or stripped of most of its entries. But the current version is the result of two RfCs (one for the general principle of changing the cryptic abbreviation + meaningless ID to a more easily understandable text link, and one for the specific layout), so it isn't "non-consensus-based" and can't be reverted based on some complaints. The last RfC gave the choicce between collapsed or uncollapsed, but opinions there were divided. I tried to demonstrate the new look on some 10,000 pages (or 1% of the uses of it) during the RfC, but this was quickly reverted. Perhaps a new RfC on whether it should be autocollapsed or not will get more input or a more clear result now the new version is actually life: but I'll let someone else start it. Just make sure, whoever wtites an RfC, that you get the history right, and don't start with claims of "non-consensus-based" and so on. Fram (talk) 07:34, 8 June 2021 (UTC)

The revert mentioned below (#Errors) has for now restored the old, less obtrusive appearance, so I can't see what it would look like at Mark Twain. As for the "new look" where opinions there were divided: the closer wrote explicitly that there is no consensus on default collapsing behaviour, so it's no surprise that people turn up here to voice concerns about this template. For the record, I agree it should be collapsed by default, at minimum it should allow a parameter |state=collapsed, like any other navbox. -- Michael Bednarek (talk) 13:09, 8 June 2021 (UTC)
I've added support for |state=collapsed in the sandbox. * Pppery * it has begun... 23:46, 11 June 2021 (UTC)
Now turn it on by default. It is horribly obtrusive otherwise. @Michael Bednarek:, re: "Anyone is free to go to the Village Pump and start a new RfC": there was no prior consensus from the RFC for making it expanded, so doing so is a non-consensus move that should be reverted. —David Eppstein (talk) 20:07, 12 June 2021 (UTC)
Support for the suggestion that it be collapsed by default. PamD 22:50, 12 June 2021 (UTC)
Support autocollapse by default. – Muboshgu (talk) 20:19, 13 June 2021 (UTC)

Pinging the participants of the recent RfC for comments, specifically on the collapsing issue: @Francis Schonken, Pigsonthewing, ProcrastinatingReader, Mike Peel, Blueboar, Tom.Reding, JohnFromPinckney, Rhododendrites, Levivich, Guettarda, CaptainEek, Sea Ane, Aza24, Ajpolino, Pbsouthwood, and The wub: sorry to ping you over such a trivial issue, but it not be fair to ping some and not all. — Martin (MSGJ · talk) 20:49, 13 June 2021 (UTC)

  • Thanks for the ping. What happened to the "default" behavior: autocollapse when there are other templates, uncollapse when it's the only template? That still seems to me to be the best approach. Barring that, if I have to choose between autocollapse or autouncollapse, then it's autocollapse, per arguments above. Levivich 21:03, 13 June 2021 (UTC)
  • As I've said all along, the recent change was harmful (as indeed would be collapsing). It should be reverted. As for the RfC, it was not (as I pointed out when it was modified) neutral, as is required Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 21:13, 13 June 2021 (UTC)
  • Old version uncollapsed is better than the new version un/collapsed.   ~ Tom.Reding (talkdgaf)  21:15, 13 June 2021 (UTC)
  • Basically every navbox is collapsed by default, why isn't this?? Especially when it is not the only navbox on a page, it makes no sense for it to be the only one that is expanded. And we shouldn't have to go through and set the collapse perimeter manually on every page that has one. The default should be collapsed, it could be uncollapsed on specific pages if folks think that is useful. But if the default is between "in your face obtrusive big ugly template that most of our readers and editors don't know what it is" and "out of the way template that can be opened by those who know what it is", I think the latter is by far more sensible. CaptainEek Edits Ho Cap'n! 22:16, 13 June 2021 (UTC)
  • I agree with Tom.Reding, the old version uncollapsed is the far better alternative. The new version looks both obtrusive and messy. However, collapsing it would be even worse. --Bjerrebæk (talk) 02:41, 14 June 2021 (UTC)
  • +1 for autocollapse behaviour. -- Michael Bednarek (talk) 03:46, 14 June 2021 (UTC)
  • Autocollapse per Levich's sensible reasoning. Aza24 (talk) 03:56, 14 June 2021 (UTC)
  • Autocollapse by default, per Levivich and CaptainEek, · · · Peter Southwood (talk): 05:29, 14 June 2021 (UTC)
  • Voicing my support for autocollapsing (in all cases, not just if present with other navboxes), if the current version is what we're stuck with. In its present form, {{Authority control}} elevates trivial information to the same—or more prominent—level as actual encyclopedic content in navigation templates. It's akin to spray-painting the warranty information or mechanic's address across the hood of a brand new car: useful to some if needed, but needlessly prominent, distracting, and just plain tacky. --Animalparty! (talk) 05:34, 14 June 2021 (UTC)
  • I've already given my thoughts on this page, but to summarise: the new version is not an improvement over the previous version, particularly for the amount of space it uses, but collapsing it by default would make it even worse as it hides it out of the way of people that would find it useful. Mike Peel (talk) 06:49, 14 June 2021 (UTC)
    • Collapsing it would make it no less visible than the previous form, and would only require one more click for those people to uncollapse. If you think that the readers who want this are incapable of making that one click, I think you need better faith in the technical ability of readers who are technically savvy enough to want to use this at all. —David Eppstein (talk) 07:16, 14 June 2021 (UTC)
      • @David Eppstein: It would make it much less visible than the previous form. No longer would you be able to glance at it and see what content it contains. Nor would you easily know if it exists on the page (many pages still don't have it), particularly as it would blend in with navigation templates. Plus, I keep clicking on the headers of collapsed templates expecting them to expand, and only then remember it's the tiny 'show' link on the right, so it's often not just one click. Mike Peel (talk) 07:27, 14 June 2021 (UTC)
  • Collapse it. ProcrastinatingReader (talk) 07:19, 14 June 2021 (UTC)
Pppery, which sandbox did you add it to? (I don't see here that you've recently edited any sandbox). If you'll point me to it, or just tell me exactly what is needed (and for the autocollapse default also), I'll add it to the page – I'm satisfied that there's sufficient consensus here for both, in the short term at least. Or of course you could do it. Caveat: this is written in a language I don't speak, but that's just nothing compared to the extent to which the page it invokes is beyond me. If there are any Luans reading this, please feel free to step in! Justlettersandnumbers (talk) 16:13, 14 June 2021 (UTC)
@Justlettersandnumbers: I've updated the module to follow standard navbox collapsing rules by default (autocollapse unless it's the only footer template), which seems to be what most people wanted.

For what it's worth, the sandbox edit I was referencing was Special:Diff/1028116858, which didn't show up on the GUC tool since it only shows the latest 20 edits to each wiki, and I've made several hundred edits since then. * Pppery * it has begun... 16:45, 14 June 2021 (UTC)

OK, great, thank you! Seems to be working as expected, many thanks! Justlettersandnumbers (talk) 16:58, 14 June 2021 (UTC)
@Pppery: Why have you made this change without consensus? Mike Peel (talk) 17:08, 14 June 2021 (UTC)
Because I saw a consensus to do it, and it was clear from my reading of Justlettersandnumbers' comments that they would have done it themselves if they were more familiar with Lua coding. The sole reason I got involved in this in the first place was to remedy the situation in which changes were proposed and got consensus but no one had the technical skill to implement them. In this section, there are twelve people (Randy Kryn, David Eppstein, PamD, Michael Bednarek, Muboshgu, Levivich, CaptainEek, Aza24, Peter Southwood, Animalparty, ProcrastinatingReader, and myself) explicitly supporting collapsing, and two (Mike Peel, Bjerrebæk) opposing it. It's rather impressive to claim that's not a consensus. * Pppery * it has begun... 17:15, 14 June 2021 (UTC)
Mike Peel, please see above where I wrote "I'm satisfied that there's sufficient consensus here for both, in the short term at least", and invited Pppery to make the change. I'd have tried to do that myself if the thing hadn't been written in Lua. So any blame for the (mis)reading of the discussion should be directed at me, while any thanks should of course be sent in Ppperys' direction. And by the way, I'm still satisfied that there was consensus for both changes, despite your opposition. Justlettersandnumbers (talk) 17:25, 14 June 2021 (UTC)
@Pppery: Consensus isn't just counting (which apparently you can't do anyway, you missed Tom and Andy at least), and the RfC closed with no consensus for collapsing the template. The template display has been getting worse and worse, as I predicted right from the start of this, sigh. Mike Peel (talk) 18:48, 14 June 2021 (UTC)
Mike Peel, isn't this whole business difficult enough for everyone without adding personal remarks to the mix (you might like to strike that one, perhaps?)? As above, it was I and not Pppery who assessed the consensus in this discussion, and then asked Pppery to make the changes because they were (and are) beyond my embarrassingly limited technical abilities. I didn't take into account the opinions of either Tom.Reding or Andy Mabbett because both expressed a wish to return to the previous version of the template but no preference for collapsing this version or not; I did take account of your opinion and that of Bjerrebæk. If I've made a mistake here please take it up with me and no-one else. Thank you, Justlettersandnumbers (talk) 20:10, 14 June 2021 (UTC)
I've given up. Have fun continuing to trash this template. I'm not watching it any more. Mike Peel (talk) 20:22, 14 June 2021 (UTC)
I wrote the recent change was harmful (as indeed would be collapsing). I'm not clear what part of that is not clear to you, much less why you would see it as anything other than strong opposition to collapsing the template. Andy Mabbett (Pigsonthewing); Talk to Andy; Andy's edits 20:53, 14 June 2021 (UTC)
I still think this should autocollapse by default rather than only in the presence of navboxes, as it is not itself a navbox. As it is, I am now looking forward to going through hundreds of recently created articles manually adding |state=collapsed to each one. What fun. —David Eppstein (talk) 17:39, 14 June 2021 (UTC)
Also, the new autocollapse seems to work inconsistently, not hiding the material when there is only one line (example: [1]) and not collapsing the template at all nor providing any "hide" button to collapse it in some cases (example: [2]). —David Eppstein (talk) 17:47, 14 June 2021 (UTC)
Re templates on one line, is there some benefit to collapsing these? The "Authority control: show" line would take up just as much space as the template does currently, so I don't see the point. Using a horizontal rather than a vertical format (and thus no logical place to put a show button) when there are a small number of identifiers was originally Jonesey95's idea in Template talk:Authority control/Archive 11#The new version takes up too much space, but it shouldn't have triggered in your specific example and only was due to another bug I've fixed in the sandbox. Jonesey95, is that suggestion still relevant now that the template autocollapses by default, or should it be removed? (Regardless, I've changed the sandbox so that an explicit |state=collapsed overrides that change and forces the template to the vertical navbox format) * Pppery * it has begun... 18:47, 14 June 2021 (UTC)
David Eppstein, I agree that it would be preferable if the default state were 'collapse' – I've already manually added one collapse parameter (here) and will doubtless spend time adding more. But in this climate I don't see any way to achieve that other than proposing it here and hoping for some support. Regards, Justlettersandnumbers (talk) 20:16, 14 June 2021 (UTC)
Isn't "proposing it here and hoping for some support" exactly what this exact discussion thread already is? And there appears to be plenty of support. I don't see why each successive discussion keeps getting reinterpreted in the most expansive way possible. —David Eppstein (talk) 20:29, 14 June 2021 (UTC)
When I voted for collapse, I meant collapse it (by default). In all circumstances. No qualifiers. I dunno why it's taken 4-5 discussions just to constantly reaffirm the principles expressed in the first RfC but here we are. ProcrastinatingReader (talk) 21:11, 14 June 2021 (UTC)
Responding to Pppery's question above: If the template has a |state= option, it should honor it in all multi-line cases, so my suggestion to remove the header (which also apparently disables collapsibility) for vertically short versions is probably no longer valid. I don't see the benefit of collapsing a single-line version, however. – Jonesey95 (talk) 21:25, 14 June 2021 (UTC)
OK, I've removed that code from the sandbox. The current sandbox logic is, if all identifiers are in one section, to show just that one section on one line (unless the labels are long enough to make it take up more than one line), ignoring |state=, and otherwise to display a header with a show/hide button, currently defaulting to autocollapse (collapse unless it is the only collapsible thing on the page). Re whether it should collapse in all cases (default to |state=collapse), I have no strong opinion. * Pppery * it has begun... 21:43, 14 June 2021 (UTC)
See Template:Authority control/testcases#Compare for comparisons of the live template (top of each pair) with the sandbox (bottom of each pair). – Jonesey95 (talk) 23:53, 14 June 2021 (UTC)
To test this properly we also need to check what they do in their uncollapsed version (if there are no other collapsible things on the page. Is there any way to simulate that on the testcases page? — Martin (MSGJ · talk) 09:53, 15 June 2021 (UTC)
Remove all but one instance of the template and preview the edit? (Nothing very interesting will happen, though, you'll end up in literally the exact same state as if you had clicked the "show" button) * Pppery * it has begun... 18:55, 15 June 2021 (UTC)


I would have left this discussion run a little longer. It was implemented less than 24 hours after I pinged the RfC participants and it would not have hurt to wait a few days before making the change.

A couple of queries:

  • On Abacus, the template is collapsed but why, because there are no other navboxes on that article?
  • On Alkane, the template is not collapsed yet it takes up three rows.

There does seem to be some inconsistent behaviour — Martin (MSGJ · talk) 20:12, 14 June 2021 (UTC)

Re Abacus, I have no idea, but it has nothing to do with this template, since if I edit the whole page, and replace {{authority control}} with a different navbox, and preview the change, the replacement navbox also shows collapsed. Re Alkane, see my reply to David Eppstein above, the collapsibility code I wrote only kicks in if the template shows a navbox header, which currently happens when there are four or more identifiers (not counting WorldCat due to a bug I've already fixed in the sandbox). I agree this is weird and nonstandard, and should have realized that when I deployed the change. * Pppery * it has begun... 20:28, 14 June 2021 (UTC)
The AC template on Abacus is autocollapsed because it is one of two elements on the page with the "collapsible" property. The other is {{Infobox Chinese}}. That's how autocollapse works; the same problem happens when {{multiple issues}} is present in an article with just one navbox. As Pppery says, if you swap a random autocollapsible navbox, like {{MTV Movie Award for Best Fight}}, for the AC template, you will see that navbox autocollapsed as well. – Jonesey95 (talk) 21:30, 14 June 2021 (UTC)
It is false that collapsability currently kicks in when there are four or more identifiers. Lori Lamel is still showing an uncollapsed and uncollapsable four-line Big
. —David Eppstein (talk) 23:47, 14 June 2021 (UTC)
Christ, the box is longer than the longest section on the article. This template is really the case study for why a template should never be deployed to 2 million articles ever again without clear, broad community consensus at VPR in advance. Textbook WP:FAIT. ProcSock (talk) 00:30, 15 June 2021 (UTC)
I've always thought adding this template to bios was mandatory because the template doc says this template should be added to all biographies. I realize now I've probably been taking that too literally. Levivich 00:37, 15 June 2021 (UTC)
@Levivich: I add it to bios I create because otherwise the article will turn up on my watchlist when someone else adds it. PamD 04:45, 15 June 2021 (UTC)
I've now deployed the code from the sandbox, so Lori Lamel's authority control template collapses like it should. * Pppery * it has begun... 20:41, 22 June 2021 (UTC)

Broken identifiers

During the process of implementing the RfC, Fram discovered that the following identifiers were broken on many or all uses:

  1. Terminologia Embryologica
  2. Terminologia Histologica
  3. BALaT (Belgium)
  4. WorldCat (via VIAF)

Is there any objection to removing these from the module? * Pppery * it has begun... 14:05, 7 June 2021 (UTC)

For Terminologia Embryologica and Terminologia Histologica, the content is still available and the URL could have been fixed (Template:TerminologiaEmbryologica and Template:TerminologiaHistologica provide working links to the same resource), however I still propose removing them since every article I checked includes the same information in the infobox so I don't see the benefit of repeating them in the authority contol template at the bottom. I've removed those two and BALat from the sandbox, and also implemented Template talk:Authority control/Archive 11#Researcher ID leads to Publons? and fixed an edge case in the implementation of #Duplicate Poland national library identiifers, which wasn't working properly when the |NLP= ID was passed as a parameter rather than via Wikidata. I plan to copy this code to the live module in a week (since there is a standard one-week delay for adding identifiers, there should be one for removing them as well).
WorldCat (via VIAF) was previously removed for the same reason in Special:Diff/949779734, and then re-added in Special:Diff/959383148, so I am choosing not to re-remove it in this proposal to avoid it getting caught up in unnecessary drama. * Pppery * it has begun... 23:46, 11 June 2021 (UTC)
Strong objection to removing WorldCat. WorldCat is one of the most practical links in what is otherwise mostly a wastebin of data of use to a handful of librarians in the world, and robots. Worldcat can actually help people find sources by and about the subject, and locate copies of the work in libraries. Using George Washington as an example, why should we remove WorldCat with all of its value but keep links to BIBSYS or FAST or Vatican? Yes sometimes the WorldCat link for a particular subject doesn't work. Oh well, that's what happens when data is automatically sucked from Wikidata. It's bad enough this wastebin of data is taking up 3-5 times the page space as before, why make it less practical for users? --Animalparty! (talk) 03:00, 12 June 2021 (UTC)
I was not proposing removing WorldCat, only the other 3 broken identifiers. Also, George Washington uses a "WorldCat" link, which is a separate thing from the "WorldCat (via VIAF)" link I initially considered removing then decided against, so wouldn't have been impacted anyway. * Pppery * it has begun... 03:04, 12 June 2021 (UTC)
WorldCat (via VIAF) has been having some problems for ~the past 2 months, at least from what editors have reported on my talkpage (1, 2). WorldCat itself is online, but via-VIAF is not, and it's not obvious to me why. There's nothing on worldcat.org explaining it, and I've tried a few incrementally-different URL variants to see if it was/n't a small/guessable URL update at fault, to no avail. It would be nice to know if this is a permanent or temporary issue. At first it appeared transient, but after weeks of 404s (I check a random article every few days), it seems to be becoming more permanent. Given its long tenure though, I think it's worth giving them the benefit of the doubt, and a little extra time before removal, say, by July.   ~ Tom.Reding (talkdgaf)  11:35, 13 June 2021 (UTC)
Makes sense to me. * Pppery * it has begun... 13:17, 13 June 2021 (UTC)
@Pppery: Something has happened to Module:Pages with authority control identifiers to put a whole bunch (technical term meaning "almost 90") of cats (such as Category:Miscellaneous pages with VIAF identifiers) inside themselves, they are listed at Wikipedia:Database reports/Self-categorized categories. --Redrose64 🌹 (talk) 09:09, 16 June 2021 (UTC)
Fixed Oops, I should have realized that would happen when I wrote that code. * Pppery * it has begun... 13:46, 16 June 2021 (UTC)

ISNI (and others) shows as a linked 1, instead of the full number. Not sure what's going on.--Auric talk 14:58, 21 June 2021 (UTC)

Not a bug, a deliberate way of including a wikilink to ISNI while including the URL somwewhere and consistently showing the redesign. This specific layout was discussed at Template talk:Authority control/Archive 11#Taking out all the wikilinks doesn't seem like improving user-friendliness * Pppery * it has begun... 15:36, 21 June 2021 (UTC)

 Removed TE, TH, and BALAT, since no one objected to removing any of them. I didn't end up doing the Publons thing, since it only affects one page so I decided it wasn't worth adding. * Pppery * it has begun... 20:41, 22 June 2021 (UTC)

@Tom (LT): notifying the TE & TH proposer, as should be the norm. I couldn't find who proposed BALaT (I searched the archives for "P3293", "BALaT", & "Belgian Art").   ~ Tom.Reding (talkdgaf)  11:57, 23 June 2021 (UTC)
Sorry, I should have thought to do that, but it didn't occur to me. * Pppery * it has begun... 13:28, 23 June 2021 (UTC)

Add support for P4613 Encyclopedia of Modern Ukraine ID

Online national encyclopedia of modern Ukraine. The ID is linked to 4,394 articles in en.Wikipedia.

 —Michael Z. 15:58, 22 June 2021 (UTC)

Disabling {{TPER}} since there's nothing for a Template Editor to do right now; there's a standard one-week delay for adding identifiers (per the header), so someone will get to this in a week. * Pppery * it has begun... 20:41, 22 June 2021 (UTC)
Re-enabling TPER since there were no objections in a week. ネイ (talk) 14:02, 5 July 2021 (UTC)
 Not done: According to the page's protection level you should be able to edit the page yourself. If you seem to be unable to, please reopen the request with further details. Elli (talk | contribs) 22:40, 8 July 2021 (UTC)
I will leave it to Michael Z then, since I do not have template editor rights. ネイ (talk) 08:37, 9 July 2021 (UTC)
Thanks. I will report back if I encounter any issues. —Michael Z. 13:49, 9 July 2021 (UTC)
 Done   ~ Tom.Reding (talkdgaf)  13:13, 15 August 2021 (UTC)

User:UBX/Authority control

{{AC}}This user uses the authority control template and the data it provides.

I created a userbox for self-identifying as an AC user. This could be helpful in future discussions as a way of pinging interested parties who may not check their watchlist regularly.   ~ Tom.Reding (talkdgaf)  14:44, 15 August 2021 (UTC)

Relying on WD but rendering only one statement ?

I was looking at Agha (title).

The template is used w/o parametrer, so as to rely on Wikidata. But WD has identifiers from five databases, and the template as rendered on Agha (title) only shows one. Why is that? trespassers william (talk) 18:45, 27 August 2021 (UTC)

@Danny lost: that is b/c Encyclopædia Britannica Online ID (P1417), Encyclopédie berbère article ID (P9198), Freebase ID (P646), & Gran Enciclopèdia Catalana ID (P1296) don't exist in the AC template, either b/c they haven't been proposed here yet, or they were & then rejected.   ~ Tom.Reding (talkdgaf)  17:17, 29 August 2021 (UTC)
@Tom.Reding: Thanks. trespassers william (talk) 20:15, 29 August 2021 (UTC)

The new design forces all navboxes to autocollapse

The horrible new design of the authority control template makes it behave like a navbox (instead of a separate template group, as before), and it now forces navboxes to autocollapse, even in articles with only one proper navbox where the uncollapsed state would be preferable. Before the changes an article with one navbox plus the authority control template would look like this Norwegian article no:Gudmund Restad (with a navbox of Norwegian finance ministers and authority control below), in other words endlessly more reader- and editor-friendly. --Bjerrebæk (talk) 20:23, 31 August 2021 (UTC)

@Bjerrebæk: you may thank the committee.   ~ Tom.Reding (talkdgaf)  20:59, 31 August 2021 (UTC)
@Bjerrebæk: (and anyone else interested in the size of the new AC): I've created the following tracking cats to provide everyone with more information:
Incidentally, this also helps catch unrecognized parameter values.   ~ Tom.Reding (talkdgaf)  14:26, 5 September 2021 (UTC)

Base Léonore links broken

Changes made to the Base Léonore website (which holds details of members of the French Legion d'Honnneur) mean that links generated by this template no longer work.

People are identified via a reference number (a "cote" in the terminology of the site) in the form LH/1924/18. This is converted into a link like this: http://www.culture.gouv.fr/public/mistral/leonore_fr?ACTION=CHERCHER&FIELD_1=COTE&VALUE_1=LH%2F1924%2F18 . This link now goes nowhere.

The page it now needs to go to doesn't include the cote at all, it's this one: http://www.leonore.archives-nationales.culture.gouv.fr/ui/notice/269341 . There's no way to find that numeric id from the cote, other than going to the site, entering it into the search box, and seeing what you find. If you do that, you need to be aware that they've added an extra slash to it, so it's now LH//1924/18. Cotes in the old style come back with no hits.

I'm not sure what the best way forward is. Léonore are actually using wikidata on their site, so they might be amenable to putting something in place to facilitate linking. I've no idea how to make that approach - it probably should be done by French wikipedia.

In the short term, Léonore cotes should probably be excluded from authority control. There's also a template dedicated to linking there that will need to be fixed - I've started a discussion at Template talk:Base Léonore

What do you think? Chuntuk (talk) 07:31, 18 August 2021 (UTC)

 Updated. Do you know what the Bis & Ter suffixes mean?   ~ Tom.Reding (talkdgaf)  11:13, 18 August 2021 (UTC)
bis can mean various things associated with the number two. From twice to encore (playing/singing again) to a suffix of a house number in an address or a second revision of a document. ter is used much the same for the number three. -- Michael Bednarek (talk) 12:35, 18 August 2021 (UTC)
Thanks, @Tom.Reding:, I've found this discussion (in French) on this issue at wikidata: http://www.wikidata.org/wiki/Topic:Wbisfmslwlmzba5u . Looks like the linking issue might be fixed one day, but the change in cote format is here to stay. I've made changes to Template:Base Léonore which use the id parameter to construct a working URL, but there are about 70 articles that need updating with revised cotes. Chuntuk (talk) 09:30, 19 August 2021 (UTC)
There are 500+ articles in Category:Articles with faulty Léonore identifiers (4,224). The one I checked, Sarah Monod, had a valid identifier but no link, and a red error message. Are the category and error message related to this discussion? – Jonesey95 (talk) 20:26, 30 August 2021 (UTC)
@Jonesey95: yes. The format as a regular expression (P1793) for Léonore ID (P640) was updated, but the values on many of the pages have not yet been updated (current faulty count = 4,237).
@Nono314: (d:User:Nono314): I see you added some of the original Léonore ID values years ago. Would you be willing to update those that need it? Most, if not all, of the updates required are simply adding a second "/" after "LH/" like so: LH/110/91LH//110/91.   ~ Tom.Reding (talkdgaf)  14:39, 5 September 2021 (UTC)
Indeed, back in the days, I sent the data to Magnus for uploading in Min'n'Match and did quite a part of the matching... I had also seen the discussion on Wikidata's French project chat (linked above) when it started. It has been initiated by @Xavier Cailleau WMFr:, who is a member of Wikimedia France and is liaising with French National Archives (managing body of the Leonore database). The current community consensus is to wait for an official statement from their side, instead of blindly "fixing" 21K+ values (and 13K+ references) on our side, without even knowing whether this will help in the future. Therefore, I'm not willing to do anything for now. Bot operators like User:Tpt are ready to act once the situation is clearer. --Nono314 (talk) 21:23, 6 September 2021 (UTC)
More discussion of the situation (in French) here: http://fr.wikipedia.org/wiki/Discussion_mod%C3%A8le:Base_L%C3%A9onore#Probl%C3%A8me_avec_le_lien_g%C3%A9n%C3%A9r%C3%A9_vers_la_base_L%C3%A9onore they seem hopeful that a working link will be established at some point. Chuntuk (talk) 11:19, 6 September 2021 (UTC)

Remove "Wikipedia" from "Wikipedia articles with EMU identifiers"-type categories?

Every AC ID has 2 categories that could be less-redundantly renamed to:

This would also save a small # of bytes from the expanded page (see above @ #Is there a known issue related to Authority Control template, and the Navboxes Top / Bottom format?).

Performing this would be relatively simple. After moving all of the relevant categories (leaving redirects), the entire Category:Pages with authority control information ID-tree is automated via {{Pages with authority control identifiers}} and p.docConfTable in Module:Authority control.   ~ Tom.Reding (talkdgaf)  14:17, 15 August 2021 (UTC)

 Working   ~ Tom.Reding (talkdgaf)  18:11, 29 August 2021 (UTC)
 Done - old cats emptied.   ~ Tom.Reding (talkdgaf)  10:47, 20 September 2021 (UTC)