Wikidata talk:Requests for comment/Commons links

From Wikidata
Jump to: navigation, search

Pragmatic approach[edit]

Commons has always been pragmatic, maybe it's good to do that here too:

  1. If a Wikipedia article exists and a Commons gallery: Link them to the same item
  2. If a Wikipedia article category exists and a Commons category: Link them to the same item
  3. If the previous options are not possible, a Wikipedia article exists and a Commons category exists: Link them to the same item

Thoughts? Multichill (talk) 15:09, 5 October 2013 (UTC)

(edit conflict) Sorry, I'm not following you. How does 3 differ from 2? --Avenue (talk) 16:37, 5 October 2013 (UTC)
Oops, corrected. Multichill (talk) 23:10, 5 October 2013 (UTC)
And the result will be a mess. We should link all similar way, and not like... "if, then". Well, this is only my opinion. --Stryn (talk) 16:34, 5 October 2013 (UTC)

What if a pragmatic approach would be a combination of proposals 2 and 5? --Sannita - not just another it.wiki sysop 19:35, 7 October 2013 (UTC)

imo, horribly wrong. instead, what makes sense to me:
  1. a Wikipedia article exists and a Commons category exists: Link them to the same item. (on C~, category is king. both 4 users & editors: are the primary navigational, and organisational feature)
  2. a Commons category cantains at top the zero-to-many Commons galleries, above the Commons files
  3. Wikipedia(etc) categories arent given a crap about when it comes to Commons linking. and not much about other linkings either, they'll do fine within their own wikis. as someone said, those cats are the back-end for users, and many are not giving a damn about them. (as i detailed on project page, a single interwiki link to the unified wikidata page, where they are listed inline, will do for these pages.)
(if some of (1) not, then don't do anything. (either the lonely art'l, or the lonely commonscat, will have some other relation within it's wiki to prevent it being totally orphaned w.o any inter-representation) --Aaa3-other (talk) 15:48, 10 November 2013 (UTC)

Review of closure[edit]

I got asked to look at the closure of this RFC, and the discussion. I have been kind of uninvolved so far in this particular topic, and I don't have any preference either way. I will look at any discussion posted in this section, on my talk page, on the RFC and its talk page. I will need about 24-30 hours for this. Thanks for John for the courage to close the very complicated RFC in the first place. --Denny (talk) 22:38, 18 November 2013 (UTC)

I have read the whole discussion. Here are my conclusions:
  • John F. Lewis handled the closing of the RfC extremely well, and made a good call under very hard circumstances. His addenda were courageous, and have to be admitted given the RfC. I completely follow his decision and agree with him in the way he has handled the RfC and defined a solution.
  • I am very disappointed by some of the participants in the discussions. They are extremely dismissive of hard work, without respect, but demanding respect themselves. I think that the behavior of some the participants is poisonous towards the collaborative climate of the project, even though I assume that they are doing this with the best motivation. I would ask those participants to apologize and reconsider their choice of words more in the future.
  • In order to close the discussion and move on, I suggest that we follow the result of the closure. Any new RfC to change the results of this proposal in a substantial way should not be allowed to stop the actual work on the data, and should only be regarded as binding if it receives a considerable more-than-majority support by a sufficiently large number of active community members. I.e. that means, simply complaining and opposing the resolution is not sufficient, a constructive counter-solution must be suggested and agreed on.
Thank you for your attention. --Denny (talk) 02:42, 22 November 2013 (UTC)
Denny, I don't think there was any disagreement that John F. Lewis did a good job closing the rfc, specially after Addendum 2. The main clarification needed here was to know what to do now with the outcome. According to my understanding, feel free to correct, the consequences of the closure are:
  • Option III is desirable in the long run and it will be re-examined at some point in the future
  • For now sitelinks will be arranged as described in Option II and VI (articles with galleries, categories with categories)
  • Commons category items will be created to represent Commons categories that have no equivalent in Wikipedia and interlinked as described in option VI
  • Some work needs to be undertaken by the development team to represent such structure in Commons (see option VI for more details).
Please, confirm that these are the accurate effects of the outcome of this RFC.--Micru (talk) 10:11, 22 November 2013 (UTC)
I'm grateful to both John and Denny for their closure and review of the RfC. There are two aspects of the closure that concern me, however. The first is John's revision to our notability policy ruling out the creation of items linked solely to Commons categories. (Micru, this directly contradicts your third bullet point.) This change doesn't seem to have any support, or even any discussion, in the RfC. John hasn't yet responded to requests for clarification about this on his talk page.
The second is that the description of the RfC outcome at WD:RFC makes no mention of option III. I suggest that something like this should be added: "This is a stopgap solution until multiple sitelinks from an item to pages in different namespaces are supported." --Avenue (talk) 12:49, 22 November 2013 (UTC)
I have not responded to said clarification requests as I asked Denny to reassess my whole closure. With that done, I though his reassessment had adequately address the clarification needed. If not, feel free to state clearly in questions the clarification you need. John F. Lewis (talk) 23:31, 22 November 2013 (UTC)
While I agree it was useful for Denny to review the closure, I don't feel he really provided further clarification of this point. My reading of his review commentary is that he endorsed your closure, without explaining it any further. He also censuring some (unspecified) RfC participants for their behaviour, and laid out appropriate procedures for anyone who might want to dispute the outcome.
I don't want to dispute the broad outcome, but I do feel it needs some clarification. (I don't believe I'm alone in this). I understand you've approved the interim adoption of option VI. But I believe this option requires the creation of many items for Commons categories (and there are comments in the RfC from other editors supporting this view). Your change to the notability policy rules out the creation of such items, so these aspects of the closure seem inconsistent to me. I presume you don't see an inconsistency here, however, so my main question is what part of my logic (or my premises) do you disagree with? Do you believe that option VI is workable without the creation of any items for Commons categories, for instance? --Avenue (talk) 10:54, 23 November 2013 (UTC)
Hi Micru, maybe I understand closure incorrectly, but I do not see consensus about your points "* For now sitelinks will be arranged as described in Option II and VI..." and "* Commons category items will be created to represent...". There are many opposes for these two points in discussion too. — Ivan A. Krestinin (talk) 18:15, 22 November 2013 (UTC)
Avenue, Ivan A. Krestinin: Then I do not understand the statement "Therefore effective when people see this, VI is successful at 100%.", since one of the implications of VI was the creation of items for categories that only exist in Commons.--Micru (talk) 18:43, 22 November 2013 (UTC)

John F. Lewis, my questions are:

  • What is the meaning of your sentence: "Therefore effective when people see this, VI is successful at 100%."? Does it mean that it is accepted (which implies creating items for Commons categories, contradicting your edition of WD:N), not accepted, or something in between?
  • Which Commons pages should be linked from Wikidata items that only have sitelinks to Wikipedia articles?
  • Which Commons pages should be linked from Wikidata items that only have sitelinks to Wikipedia categories?
  • What do we do with Commons category (P373) and Commons gallery (P935)?

--Micru (talk) 10:46, 23 November 2013 (UTC)

Hi Micru,
  1. It does mean it is accepted and my wording of WD:N was indeed contradictory after a user pointed it out to me in regards to warning. The policy has been updated to reflect this thanks to GZ.
  2. Galleries.
  3. Categories.
  4. We used P373 to show a category relationship on a gallery item (articles) and we use P935 to show a gallery relationship on a category item (Wikipedia categories).
John F. Lewis (talk) 12:31, 23 November 2013 (UTC)
John F. Lewis, thanks for clearing it up!--Micru (talk) 12:41, 23 November 2013 (UTC)
Thanks for this clarification, John, but I'm sorry, this doesn't seem to clear up the main point. The current version of our notability policy still prohibits the creation of items linked solely to Commons categories, and this is still entirely inconsistent with the general understanding of option VI. (The wording change did make the prohibition clearer, but that doesn't solve the problem.)
I believe that this prohibition should therefore be removed from WD:N. Do you agree? If not, please explain why you think this prohibition is consistent with option VI. --Avenue (talk) 13:28, 23 November 2013 (UTC)
Okay, since there is still no response (and I raised this concern almost a week ago) I will be more forthright. I do not believe this addition to WD:N had any support in the RfC (or was even discussed), so I do not see this as a legitimate part of the RfC closure. Several of us believe it is also highly inconsistent with the interim approach approved in the closure (option VI). I therefore intend to remove the new clause prohibiting items with only Commons category sitelinks from WD:N if no one objects within the next 48 hours. --Avenue (talk) 09:44, 25 November 2013 (UTC)
Well I object. The wording now is fine in that commons links may only be used in category items. Also whether it was directly discussed is different to whether this change is appropriate. The said change is appropriate inline with discussion had therefore there is consensus. Also problems can not be solved with acting against any form of consensus. John F. Lewis (talk) 16:42, 25 November 2013 (UTC)

@John F. Lewis: I would expect any positive position in a closure to have been discussed, in any RFC at any wiki. Without such, you cannot (bolded for lots of emphasis) show community support for a decision on your part. You seem to have created policy by fiat, because you did not explain then and you still do not explain now why this is an "appropriate inline with discussion". More fundamentally though, you do not seem to understand that positions 2/6 would require the creation of (some of) these category items because both of the options are predicated on the notion that we can link from item-about-mainspace-topic A to item-about-categoryspace-topic B using category's main topic (P301) and topic's main category (P910).

But perhaps you still have not worded yourself clearly though on WD:N, which is where a potential communication breakdown may be occurring. Maybe you are banning the creation of category items for which no mainspace items exist. Example:

  1. horse (Q726) exists and Category:Horses (Q7145036) exist. Add the Commons category on horses to Category:Horses (Q7145036) and link the two using category's main topic (P301) and topic's main category (P910).
  2. USS Wyoming (Q1715770) exists but a "category item" does not exist. Create a "category item" for commons:Category:USS_Wyoming_(SSBN-742) (and link as before).
  3. Neither a "main topic item" exists nor a "category item" exist (or more simply, a "main topic item" does not exist) for the case Commons:Category:Veterinary_medicine_for_horses. Create no items whatsoever.

Are you attempting only to ban #3? As worded, WD:N bans #2 as well, and that is certainly a failure on your part, to some extent or another.

Of course, that still does not explain why we should not support all of the Commons category pages, and you cannot seem to put a finger on why you think that should be the case. Maybe you were trying to create a compromise position where none existed in discussion. Who knows? That's not your job when closing, however. If there are compromise positions, they should and would have been discussed.

Either admit your mistake and change your decision, choose not to figure out why you think this should be the case from your reading and then face the community's wrath, or figure out why you think this should be the case... and face the community's wrath. :) --Izno (talk) 00:42, 26 November 2013 (UTC)

To be honest this discussion is dragging too much now but anyway, I have made a final update to WD:N to resolve said misunderstanding. In response to what you said, I do not see how I acted inappropriately by creating my own consensus, if you need to do something to do something else which has consensus, you do it. You don't wait another few weeks for people to give that a consensus because this is a waste of time. Also I did change my decision, twice. The reason why not three times? Because I feel the current decision is supported and is what the community wants. Whether two people disagree and them make a wild situation out of it is their opinion and we should not take time away from what we are doing to deal with it. If you feel the community has wrath, then please do imitate a discussion over me because I honestly do not see a problem with my closure. The decision to input VI is simple due to technical limitations be addressed by WMDE when they have time. John F. Lewis (talk) 07:44, 26 November 2013 (UTC)
Thank you for trying to resolve the problem. Your update leaves WD:N saying "In addition, an item with only a sitelink to a category page in Wikimedia Commons is not allowed on main article items." Taken literally, this seems to be talking about items with sitelinks to other items, which this RfC didn't cover and isn't implemented yet anyway AFAIK. So I'm not clear what you mean. Is the following close to what you meant? "Items with sitelinks to mainspace articles should not also have sitelinks to category pages at Wikimedia Commons." That would seem unobjectionable to me (except it probably wouldn't belong at WD:N at all). --Avenue (talk) 12:00, 26 November 2013 (UTC)
@John F. Lewis: Could you answer the above question?

But aside from that, maybe you're misunderstanding the direction I'm coming from. Let me go line by line: "I do not see how I acted inappropriately by creating my own consensus": Options 2 and 6 require the use of items which solely contain Commons links. The way WD:N was written previously would have banned those items. WD:N is now silent on the point, besides the more general "if it is a link, it goes in Wikidata", which is fine, but the note about Commons links is still ambiguous as to the decision.

"Whether two people disagree and them make a wild situation out of it is their opinion" – I don't disagree with the decision! I disagreed with the way you seemed to write it! And please do not inject hyperbole nor ad hominem. Our concerns are legitimate—we're seeking clarity.

"If you feel the community has wrath, then please do imitate a discussion over me" – My personal observation is that you don't take the time that you really need to to close RFCs or provide the necessary explanation (or like-discussions), and this has caused ambiguity and pain for a large number of users.... I suspect that there are others who feel the same. You might improve the closing of future discussions by having someone do a readthrough of your closures. --Izno (talk) 15:02, 1 December 2013 (UTC)

Hi John F. Lewis, could you please clarify your decision to accept variant II+VI and ignore negative sides and oppose comments of this variant? — Ivan A. Krestinin (talk) 19:12, 23 November 2013 (UTC)
What oppose comments? A single user going 'This is wrong in my opinion' is not enough for a whole RfC subsection to be declined. I based the whole conclusion on VI. I considered II and didn't feel consensus was there, therefore II was ignored further. Whether VI is a variant of II does not matter. VI is a proposal in its own rights therefore all other comments were ignored. John F. Lewis (talk) 19:46, 23 November 2013 (UTC)
As I comment in RfC: short discussed IV is very similar to long discussed II, so arguments that was discussed in II must be applied to VI too. Discussion of VI was not only short, but contains delusion too: [1]. Closure in ruwiki usually contain and is based on discussed arguments analysis. But your closure is simple vote counting. This is looked strange. Is there NOTDEMOCRACY principle in Wikidata? Or all problems are solved using voting in Wikidata? — Ivan A. Krestinin (talk) 20:43, 23 November 2013 (UTC)
It was an argument based closure. If not, VI, would not have been 100% successful and III would never have been recommended. Also argument based closures vs vote based closures are always a fine fine line. Sometimes we have to defer to vote closures and when possible, we use discussion closures. The reason why vote closures play a factor in all closures is a simple proposal with 1 great support again 19 opposes, can be successful using a argument closure however it is not what the community wants. John F. Lewis (talk) 22:28, 23 November 2013 (UTC)
So we select III as major goal, we have the next ways: I → III, II+VI → III. In your closure II+VI → III is selected. What benefits have this variant? Why these benefits are more important then negative sides of this variant? — Ivan A. Krestinin (talk) 07:32, 25 November 2013 (UTC)

What is the point ?[edit]

I totally disagree with the result of this RFC. It serves no purpose, it adds no value. My arguments can be found here. Thanks, GerardM (talk) 05:59, 25 November 2013 (UTC)

The point was to find a solution for interwiki-linking with Commons until the new "structured data Commons" is ready. We don't know yet if categories and galleries will be still relevant when that is done, therefore this RFC can be considered indeed a waste of time. OTOH, categories may remain the way they are now, in that case it would make sense to start discussing it now before development begins. Your blog entry lacks in detail regarding what to do with categories that have no Wikidata item (for instance, commons:Category:Disobedient horses). Also ommitted is how to convey to the reader more granularity starting from a general term like "horse" (this is the job that categories are doing now). --Micru (talk) 10:07, 25 November 2013 (UTC)

Preparing bug report for Option III[edit]

I was thinking what to write in the bug report for Option III, which seems to be the most wanted option. Let's see if I have it right:

  • Create a new sitelink group named "Wikipedia [topic's main] categories linked to this item"
  • Allow two links in "Wikimedia Commons pages linked to this item", one for the category and another for the gallery

Anything else?--Micru (talk) 20:30, 25 November 2013 (UTC)

Such a feature already has a bug report somewhere plus this RfC is on her todo so filing a new bug is just creating a duplicate one and increasing whatever Bugzilla work she has to do. John F. Lewis (talk) 20:59, 25 November 2013 (UTC)
What is the bugzilla number? I cannot find it.--Micru (talk) 21:18, 25 November 2013 (UTC)
Something to do with allowing more than one site link per site. Plus either way, Lydia knows of this and will be addressing it (planning etc) probably after the quantities datatype is released. John F. Lewis (talk) 21:22, 25 November 2013 (UTC)
Searching bugzilla for 'sitelink' returns 42 bugs, none of which seem to address option III. (The closest thing I can find is bug 55090, but that is about a wikisource issue, and isn't even a confirmed problem.) Other searches fare no better. Better to mistakenly add a duplicate bug (which can simply be marked as a duplicate, once the original one is located) than to risk having no bug report at all, particularly for something this vital.
The point isn't whether Lydia knows about the problem; it's whether the problem's being tracked properly and transparently. It won't reduce her workload if we have to keep asking her personally about progress, instead of being able to simply look it up on bugzilla. --Avenue (talk) 22:32, 25 November 2013 (UTC)