Template talk:Artwork

From Wikimedia Commons, the free media repository
Jump to: navigation, search
Info non-talk.svg Template:Artwork has been protected indefinitely because it is a highly-used or visible template. Use {{Edit request}} on this page to request an edit.
Please test any changes in the template's /sandbox or /testcases subpages, or in a user subpage, and consider discussing changes at the talk page before implementing them.

Don’t mark that source is “no source”[edit]


Now the template adds the fileinfotpl_src id to the source field, even if it contains only {{Source missing}}. As in {{information}}, this should be added only if the field displays source indeed. --Tacsipacsi (talk) 16:36, 3 March 2016 (UTC)

✓ Done --Jarekt (talk) 18:40, 3 March 2016 (UTC)

Add tracker category for accession number[edit]

As you might know I'm working on matching images of paintings here on Commons with items on Wikidata, see User:Multichill/Same image without Wikidata and linked pages. To make the matches better it would be extremely convenient if the accession number is tracked in the database. I want to propose to add the (hidden) tracker category Category:Artworks with known accession number. This is also how the monument templates work. @Jarekt:. Multichill (talk) 08:05, 21 August 2016 (UTC)

I am traveling at the moment, so unless someone beats me to it, I will do it when I have access to a computer. Ping me again if I forget. --Jarekt (talk) 00:43, 22 August 2016 (UTC)
Tested it in the sandbox and enabled it. Let's wait now. Multichill (talk) 16:11, 27 August 2016 (UTC)
@Jarekt: the category is slowly filling. I updated my bot to use this new data. The suggestions on User:Multichill/Same image without Wikidata are much better now. Multichill (talk) 17:44, 28 August 2016 (UTC)
@Multichill:There is some problem with File:Diorite-Reims-Damour-Post.jpg as the page shows the "[[Category:Artworks with known accession number| MHNT PRE.2009.0.215.1]]" text. I can not look at this at the moment, but will try latter. --Jarekt (talk) 13:17, 9 September 2016 (UTC)
Blame {{object photo}} and {{Category definition: Object}}. Keep running into those. I would love to deprecate those. Multichill (talk) 13:21, 9 September 2016 (UTC)
Yes they are odd. I think they are result of failed proposal for "Artwork" namespace run in a similar manner as "Creator" and "Institution" namespaces(See here.) Once we rewrite {{Artwork}} to pull most of the data from Wikidata, we might be able to implement {{Object photo}} so it takes just q-code instead of category name. --Jarekt (talk) 15:27, 9 September 2016 (UTC)

Multichill, This change broke infobox for File:Józef Piłsudski - Sprawozdanie z konferencji CKR - 701-001-162-001.pdf and probably all the other files from that institution, since accession number is a table now. I think we need to revert or rewrite. --Jarekt (talk) 17:51, 5 October 2016 (UTC)

Warming up this pending thread pinging Multichill and Jarekt: It's not a problem of a few files like the ones mentioned above, there are at least many hundreds that break categorising into Category:Artworks with known accession number from Category:Muséum de Toulouse: Inventory/Prehistory or from the Louvre (File:David Self Portrait.jpg for example) or from Rijksmuseum Amsterdam (File:SB 1175-De Dam met het Stadhuis in aanbouw-De Dam, naar het westen gezien, met het Stadhuis in aanbouw.jpg for example). This issue could be solved easily if the categorising were done without adding a sortkey, because sortkeys obviously must not contain links, the links break it regularly. In general I think the tracking categories are a good idea providing fast db queries, but the sortkeys lead to more trouble than benefit. The links showing up on the pages' respective ID lines are merely useful, so I suggest categorising without any sortkey. The sortkeys don't show up on the cat page anyway. --Achim (talk) 19:37, 26 January 2017 (UTC)
@Achim55: you're talking about trouble versus benefit, but you don't mention the benefits. The whole point is to have the sortkey available so we can match things. Over 70.000 files have been matched based on this. I agree the messed up output is not good. Let's do it like this
  1. If the file doesn't have an accession number -> Don't add a category (like now)
  2. If the file doesn't have a Wikidata link, but does have a accession number -> Put in Category:Artworks with known accession number with accession number as sortkey (like now)
  3. If the file does have a Wikidata link and have a accession number -> Put in Category:Artworks with known accession number and use inventory number (P217) of the item (if available).
That probably covers most problematic files. Other option is to switch to LUA. That should be done anyway in the long run. Multichill (talk) 20:17, 26 January 2017 (UTC)
Multichill, well, looking at the category I can't of course see whatsoever is done in the background using the sortkey data. Can't we simply remove the [] brackets while assigning the data to the sortkey value making the links appear as simple text there? Or remove the links at all as they are not part of the number? --Achim (talk) 20:46, 26 January 2017 (UTC)
Thanks Jarek for fixing this. That seems to be a good solution. Multichill (talk) 13:52, 16 December 2017 (UTC)
@Jarekt: do you think it can be raised to say 20? On files like File:Sir Joshua Reynolds - Francis Rawdon-Hastings (1754-1826), Second Earl of Moira and First Marquess of Hastings - Google Art Project.jpg this causes the accession number to be truncated from "RCIN 407508" to "RCIN 40750". Multichill (talk) 22:09, 28 December 2017 (UTC)
Fine with me. 10 was the longest accession number I could think of while trying to prevent including some very long strings which are more likely to include html markup which was messing up templates. --Jarekt (talk) 22:16, 28 December 2017 (UTC)
Raised it from 10 to 15. I hope that is enough without causing any problems. Multichill (talk) 08:30, 29 December 2017 (UTC)

Fall back to Wikidata for Accession number[edit]

Hi, I just enabled fallback to Wikidata if the Accession number is not set. It tries to retrieve the inventory number and, if it works, shows the inventory with the message at MediaWiki:Wm-license-artwork-id-from-wikidata which includes a link to Wikidata (example: "retrieved from Wikidata"). It also puts the file in Category:Artworks with accession number from Wikidata. I tested it in the sandbox including some edge cases like multiple inventory numbers and all seemed well. If this is causing problems feel free to revert or ping me to do so. Multichill (talk) 16:32, 9 September 2016 (UTC)

I love the fallback since it exists! To make the difference less obtrusive I'd like to suggest to use the small pencil icon many Wikipedias use in Infoboxes, indicating that the data comes from Wikidata and put a hyperlink to the specific property on the item on it. That could also be might for the creator templates. --Marsupium (talk) 17:05, 23 August 2017 (UTC)

Mobile version[edit]

On a smartphone, the template of the Artist and the Institution are too wide. These tables should be collapse. Is it possible? Pyb (talk) 16:22, 21 March 2017 (UTC)

AFAIK, not even collapsible templates can be collapsed on mobile web (whereas all tables are collapsed in the Wikipedia apps), and this table should not be collapsible on desktop. However, tables have their own “scrollbars” so that not the whole page becomes wider than the screen. I think it’s not such a serious problem, the result looks fairly good. --Tacsipacsi (talk) 21:45, 21 March 2017 (UTC)

What comes first? The Chicken or the Egg? The Artwork template or the wikidata item?[edit]

I have lately uploaded files of artwork to Commons. Where I found a Wkidata-Item to the object, I added it. Where there is no Wikidata-Item, it would be nice to have one. Now I hate to do work twice. I use the description the museums provide and add them to the Artwork-Template. Is there already a way that Wikidata extracts this information to built new items, or incorporate them into existing items? Or can I built the Wikidata-Item and than get a meaningful description on Commons. I work with Lightroom and use the LrMediaWiki-Upload Tool.When adding a known Wikidata-Item, it only adds the Wikidata-Icon without further description to the Title of the image.

I shortly want to upload images from museum at home, large enough to have a unique and valuable collection, but still unbothered by the Google Artproject. I think it would be nice, if we could show them what using Wikidata could do to their collection. --Wuselig (talk) 22:38, 18 January 2018 (UTC)

I agree that at the moment proper way is fill both Wikidata item and artwork template, which is tedious and most of the work is duplicated.
At the moment Wikidata are capable to store most of the artwork-related information and it is possible to improve this template so that it can fetch these data from Wikidata. Wikidata paintings project (d:Wikidata:WikiProject sum of all paintings) is very active, so some collections and creators are very well covered (see d:Wikidata:WikiProject_sum_of_all_paintings/Top_collections and d:Wikidata:WikiProject sum of all paintings/Top creators by number of paintings).
I think it is time to start to discuss about that, I'm a big supporter of this improvement.--Jklamo (talk) 14:35, 5 February 2018 (UTC)
@Wuselig: in the past the {{creator}} template contained a lot of redundant information, but thanks to Jarek we can now just have it fetch the data from Wikidata (example). I expect this to happen with the {{Artwork}} template too and I hope Jarek will have time for that soon :-)
So in the future the work flow would be that you upload an image of an artwork that references the right Wikidata item. The {{Artwork}} template fetches all relevant info about the work of art and you only have add the data about the photograph (own work and that you took it).
I'm slowly getting started with German collections on Wikidata. Last week I worked on the Berlinische Galerie, see d:Wikidata:WikiProject sum of all paintings/Collection/Berlinische Galerie. Looks like a nice collection, I added it to the work list on Wikidata. Multichill (talk) 19:05, 5 February 2018 (UTC)
I am switching to {{Institution}} template / Module:Institution/sandbox now. I hope that great similarity between {{Creator}} and {{Institution}} will allow me to be done with that one faster. Artwork template is the next one on the list as it relies on Creator and Institution templates. --Jarekt (talk) 19:22, 5 February 2018 (UTC)
Thanks, that sounds great. So at the moment, it would be best to do the editing at Wikidata and than just add the wikidata-identifier to the uploaded image. I am still getting started with Wikidata. So excuse the question: Is there a mask in Wikidata, which I could fill out analogous to the Artwork-Template. You see I started here with my local museum. Each object qualifies as a distinct Wikidata-item. So how can I most efficiently get all the information from the museum webpage and/or the Museum label to Wikidata? --Wuselig (talk) 13:33, 6 February 2018 (UTC)
Wuselig So for time being I would be uploading metadata to both Artwork template and Wikidata. However in the long term Wikidata upload is more important because that is where the data should (and likely will) reside eventually. As for data upload to Wikidata, two tool I could not live without are QuickStatements for data upload and wikidata Query service for reading and matching your data to the content of the wikidata. I am frequent client at d:Wikidata:Request a query asking for help with queries I have problems with. There seem to be many users eager to help. --Jarekt (talk) 16:52, 6 February 2018 (UTC)

Missleading text[edit]

Hi, in the description there is the sentence: "for all other dates use {{date}} or {{other date}}". I think we should not mention the date template because (a) is not needed and (b) it is not true that date can be used for other dates. Maybe this is the reason why i found lots of wrong usages of the date template in connection with the artwork template. --Arnd (talk) 19:27, 15 March 2018 (UTC)