So I would need to manually created the Wikidata items for every painting I uploaded? Do not we have a bot doing this? Ymblanter (talk) 20:56, 9 January 2023 (UTC)
Some people import from Commons, but in my experience that's a lot of work and the data is of not very good quality. Doable for specific sets, but not for very large ones. It's generally easier to just import from the source and match up the images (or just upload better quality newer images). Multichill (talk) 21:03, 9 January 2023 (UTC)
The collection is Q2638817 and is probably machine readable, but the problem is that very few images are uploaded on Commons, and indiscriminate upload is not a good idea because many are still not in PD. I am adding the structural data when I upload images, but I guess I am not using all fields I could, and I am not sure the bot would be able to read all the data from my uploads (w:File:Adolphe Valette Castlegate Salford.jpg would be an example). Ymblanter (talk) 21:16, 9 January 2023 (UTC)
Palissy database as an alternative to Joconde database
I always wondered why some French museums don't enter any items in the Joconde database, and I found the answer yesterday. Take the Musée des Beaux-Arts de Mulhouse. Zero items on Joconde, but 188 on Palissy: [1]! In fact, the search for "painting" (tableau) alone returns 70,000 results on Palissy: [2], of which quite a number belong to public collections. Of course, there's a huge downside, which is that no inventory number is given. But apart from that, what a gold mine of data! Edelseider (talk) 11:18, 30 January 2023 (UTC)
Thanks for the tip! If it's any consolation, the RKDiamages database is far from complete for Dutch museums and this is a much smaller country. Jane023 (talk) 13:24, 30 January 2023 (UTC)
Well spotted, Multichill. That leaves a lot of paintings to be mined! Base Joconde is updated monthly and some museums (for instance Besançon) have now put almost everything there - Wikidata is terribly lagging behind. But I have the feeling that some others museums will never touch that tool, for whatever reason. Maybe a lack of personel. --Edelseider (talk) 11:20, 31 January 2023 (UTC)
@Multichill: We actually have a bit more as some have a not too optimal modelling (even title is missed here). Palissy has too parts: items just inventoried (IM ids) and those granted some protection level through heritage designation (PM ids). Only for the latter has there been a mass upload back in 2017 (we're lagging since then) and the choice was made to base it on open data dump as source, for copyright safety, but it actually included only publicly held objects. Note that a large overlap is expected as all objects with a PM id should also have an IM id. On the other side some entries actually address more than one painting... So the actual number of real paintings is probably somewhere in-between.--Nono314 (talk) 15:21, 6 February 2023 (UTC)
@Edelseider: How is this an answer to why they do not contribute? :) The answer is probably more linked to each museum's policy (for example MBA Lyon used to have a lot in Joconde but now it's almost gone). Your finding is quite unexpected as both databases are supposed to be almost disjoint in their targets, but seems quite useful especially for smaller museums who have (almost) no presence online. I guess you'll be playing with this for some time! Note, about inventory numbers, for Musée des beaux-arts de Mulhouse (Q3330211) you can find them in the scanned 80s paperwork (top of second page) that is linked in the see also section --Nono314 (talk) 15:21, 6 February 2023 (UTC)
I noticed it when the pop.culture site launched (2019). I scraped the new site and compared to what I had scraped from the old mistral interface couple years earlier and found discrepancies, main one being Lyon nuked from ~1800 to a handful. I first thought it was just not moved to the new site, but double checking, I found it had also been removed from the old one. See, almost all existing links are dead. Nono314 (talk) 20:58, 6 February 2023 (UTC)
@Edelseider: I'm sorry what? Joconde and Palissy may sometimes by accident intersect but that's not much more than than, just an accident (just like Joconde and Mérimée or Mérimée and Palissy or etc. may intersect). It is still a good source for data import on Wikidata (a lot of data has already been imported, qv. supra, more could/should be imported). As for the main reason for (not) entering data in a database, 99% of the time it boils down to "not enough ressource: time, money, tool and/or people". Cheers, VIGNERON en résidence (talk) 14:41, 7 February 2023 (UTC)
Indeed we discussed this topic couple months ago, right after you had imported Musées Nationaux Récupération (Q19013512). The current Louvre Collections site went online less than 2 years ago. I think the bulk of the import was made quite earlier, when there was only Atlas (Q7585771) around, which focused mainly on artworks on display in the museum itself. And after that there were mainly (semi-)manual creations, based mostly on what was available on Joconde or commons.
I can definitively give a hand, as I already have some scraping code available and know about some specifics about how we handle it on wikidata (eg paying attention to locations). Nono314 (talk) 20:54, 29 November 2022 (UTC)
We were young. I was not working in the Louvre at this time.
The import was made by a scrapping of Joconde database, based on the location in the Louvre. Actually a large part of Louvre's paintings are not in the Louvre. They are in others museums or official buildings. The majority of the missing paintings are not in the Louvre storage. That's why the notion of long-term loan (Q94796160) is so important for this collection. The second issue is that data from Joconde were produced in the late 90s/early 00s and that there have been new paintings acquisitions and moves since.
Yes a complementary import could be done. In the lot there would be some long-term loans paintings already on Wikidata. 9 from the Musée des Augustins (Q2711480) were thus identified last week by Nono314 (example: The Calling of St. Matthew (Q55236582). For additional information about the data, there is a JSON version of the records; the documentation is here: https://collections.louvre.fr/en/page/documentationJSON . Note that many creators have a Wikidata ID in the JSON file: "creator": [{...."wikidata": "Q49987"}].
If a complementary import is made, there will surely be a need for additional contribution, especially for finding merges to do, adding locations and images. I would gladly contribute. Best regards --Shonagon (talk) 22:45, 29 November 2022 (UTC)
Felt like working on the Louvre so made the bot to index the paintings. It's currently running on the existing paintings adding some missing data and references to existing data.
Related to that. The combination of collection and inventory number is used to identify a painting. So every painting should have an inventory number in Department of Paintings of the Louvre (Q3044768) and inventory numbers should only be used once. Currently finding some issues here:
https://w.wiki/66Gw is a list of inventory number collisions. I see mostly combined works where the inventory number should only be on the item that has the one to one relation with the work in the Louvre.
See https://w.wiki/66KY for a list of 1300+ paintings that probably have a different inventory number than what is listed in ARK. That's a lot of paintings to either check or merge.....
Just found out the bump in numbers on property dashboard this morning, and keeping up with latest events... Nice monday surprise :)
Inventory numbers are strings subject to formatting issues and can not be expected to be fully identical across sources. That was especially witnessed working on RKD_to_match/Louvre where there are even more variations. Why not stay safe with the external id when available like here?
Please do not use "Numéro dépositaire" as this would be the other museum's number (where the painting is actually located), with even more quirks involved. Looks like most paintings without Department of Paintings of the Louvre (Q3044768) numbers are just that: entries with only "Numéro dépositaire" listed (I can see also a few historical belongings, imported from RKDI, to what was then called Central Museum where Revolution's armies brought artworks they had collected).
So quite a bit of work to find and fix all these, having to wait for bot fixes also so that they won't be reimported this way --Nono314 (talk) 12:22, 12 December 2022 (UTC)
Looks like things did not go as expected with dimensions either (did something?). At least here we are on the lacking side, rather than plain wrong. We have height without width (!) as in Q115669672, apparently as a result of the bot trying to retrieve sizes by index instead of iterating, or just nothing as in Self portrait (Q115674807) as it didn't expect 3D. --Nono314 (talk) 23:02, 12 December 2022 (UTC)
Bit busy and distracted by other things so I didn't get back here yet.
I'm pretty sure library sciences teaches you (like computer science (Q21198)) to keep your identifiers (inventory numbers) consistent and unique. And that's the case for the vast majority of collections. RKD is mostly based on hand written notes and I wouldn't reference that for an example of consistency. I'm no fan of specials or exceptions because it doesn't scale and always a source of issues, but looks like no other choice here than to switch to the ARK. I have to check the other things and I'll get back to you. Multichill (talk) 18:45, 19 December 2022 (UTC)
Thanks for getting back on this. I'm pretty sure computer science (Q21198) also taught you the difference between data and format ;) Having consistent identifiers doesn't prevent different formatting, especially in different systems. Most museums started with hand-written inventories and it's not rare at all to see for example zero-padded numbers in digital collections, or optional spaces or dashes. What is more surprising is the absence of own numbers on some paintings, maybe User:Shonagon knows why?
Then you also seemed to make a number of assumptions (eg all paintings belonging to Department of Paintings of the Louvre (Q3044768)), which ended being wrongly sourced as "stated in". And, guess what, inventory numbers are only guaranteed to be unique within department, that's why those are used as collection. We offered to help with some specifics we had already experienced, that's for a reason.
I kept away from almost any fixes until now, for fear I would only make things worse in case the bot ran again. I will continue so until you confirm the status and what you can tweak on artdatabot. Nono314 (talk) 13:56, 22 December 2022 (UTC)
Hello Nono314 The thing is that some objects registered in the collections do not have inventory number. We know that the use of inventory registers is old and the use of numbers came after. For some objects, that have not remained long in the Louvre and are on deposit since a while, there is no number yet. The ten-year collections review will be the opportunity to settle these cases and as result a collection reassignment or a number. The identification of the works is generally more complex than one would imagine, for many reasons (frequent changes of titles, bad attributions, missing works, not located, contradictory documents, duplicates, ...) and especially for a central collection. With the possibility of relying heavily on computer data, which is quite recent, it has progressed a lot in recent years. These cases encountered should be resolved in the coming years. Best regards --Shonagon (talk) 20:34, 29 December 2022 (UTC)
I started doing some improvements to correct my mistakes:
Something odd going on with the query. Some results don't show up and other results I don't expect are included. I'm running asc and desc now to hopefully get everything
I looked at the current locations for the paintings. If the painting is in some other museum, that is set as (preferred) location. If the painting is on display, the room is set as (preferred) location. So far so good.
I would agree on a consolidation on Louvre Museum (Q19675), since the reserves are no more located in Louvre Palace (Q1075988) but now outside Paris, and "temporary movement" does not seem to me like something we can track. This, of course, provided there is no mention of a location elsewhere (see issues above). Some thoughts, Shonagon? Nono314 (talk) 12:35, 12 December 2022 (UTC)
As explained before, the majority of the missing paintings that have been added are not in the Louvre, neither in the museum storage (Q1191732). Without effective information on location, the statement location (P276)Louvre Museum (Q19675) / museum storage (Q1191732) is speculative and in large part wrong. The Department of Paintings of the Louvre as central collection dispatch many of the paintings in other places, museums, that are often indicated but not always, and official buildings, that are not indicated for security reasons. Some are in museum storage (Q1191732) but it is not possible, with the data exposed, to know which ones.
The is an issue on the data quality and the statements about storage without a real reference about location should me removed, especially because we are doing now a precise, really referenced and updated contribution on the locations.
By the way I recently made a major update on about 2500 items (and a check for several thousands) about the location for artworks of Louvre's collections, not only paintings, based on the actually exposed data. It is a complex and time consuming activity (and volunteer I specify, doing it on my free time). This is worthwhile because it provides very usefull and accurate data for reuse and support for other Wikimedia projects. Best regards --Shonagon (talk) 08:12, 14 December 2022 (UTC)
My reasoning was that all the paintings must have been in the Louvre at some point in time before they got spread around again so location (P276)Louvre Museum (Q19675) is correct, just not the current value. These paintings must have at least passed through the Louvre, right?
Dates for previous collections can sometimes be found in the other links. I can probably copy some of the collection qualifiers to the location, but most of it will be manual work. Completing provenance is a lot of work. Multichill (talk) 18:23, 19 December 2022 (UTC)
Yes sure. The musée Napoléon III was located in the Palais de l'Industrie (another palace destroyed in 1896) and was not a part of the Louvre neither physically nor legally. Artworks from the Campana collection were first attached to this museum, officially called "Musée Campana" (see Collection Campana, entre le musée Napoléon-III et le musée du Louvre). When this museum was closed, artworks were attached to the Louvre's collections; many artworks went to the Louvre and many were dispatched in other museums. Let's not miss the point. Not all artworks of the Louvre's collections had been physically in the Louvre (another example: cl010213684) and today a lot are not in, neither in Louvre museum storages. Best regards --Shonagon (talk) 23:54, 20 December 2022 (UTC)
Hello Multichill Editing locations and long-term loans for Louvre's collections items is on my to-do list for the next few weeks. The locations for paintings on display in Louvre has been done a few weeks ago. Yes it's bit complex, with many cases, but I would do it with pleasure, especially as this data will be reused. Best regards --Shonagon (talk) 09:15, 22 January 2023 (UTC)
Now, following the recent massive paintings upload, there are still many duplicates to find with long-term loans and images to add. It is not an easy task, because there is very often no matches with titles and numbers. It will take some time; I started the task and will stay on it for the next weeks. Thanks for the upload. It is not an easy data set but we manage and a lot has been already done with the pleasure of seeing all this enrichment. Best regards. --Shonagon (talk) 07:19, 6 February 2023 (UTC)
Hello @Shonagon, thanks for the great work! I started taking part in the duplicate hunt, beginning with Versailles and Lyon, also adding several images to the latter. I will probably also continue from time to time. The drawback of the "depot" attribute is that the corresponding collection does not blend nicely with its non-attributed counterpart after merge. Maybe some large scale fix could also be done on inv number that were mis-assigned to Louvre (those indicated as "numéro dépositaire" in the metadata). Cheers -- Nono314 (talk) 22:32, 7 February 2023 (UTC)
Hello @MultichillL'Astronome (Q544315) has left the Louvre for Louvre Abu Dhabi(Q3176133) for a 2 years(?) loan. That's why it is not in Vermeer exhibition at the Rijksmuseum :(. It would be better for next updates to add end time (P582) instead of removing the room locations. There is no really permanent locations; it changed rooms some few years ago. In other hand we have an updated data on Wikidata about locations, which is a usefull information. Best regards Shonagon (talk) 21:30, 9 February 2023 (UTC)
Paintings of the Museum of Brittany
Hi,
Just a quick message to let you know that I imported - with OpenRefine - all the 215 missing paintings from the Museum of Brittany (Q3329701) (there was only 2 before) see the list on https://w.wiki/68Zs.
I am aware that there is still some little problems (data missing, even strange/wrong data sometimes, maybe also other problems I didn't spot yet, feel free to point them to me) but I preferred to do the import first (and before the end of the year), I'll correct these "details" manually afterward (and I'll also import some of the photos on Commons ; then I'll focus again on the others objects of this museum).
For the description, I put it as a placeholder and I'm open to any suggestion. I could add the "rather obscure" author (when it is known) but beside that, not sure what would be good. Is it a good idea and would it be enough, do you see anything else?
Actually the reference is an internal file with more data than the public portal http://www.collections.musee-bretagne.fr (when it is, I'll add Museum of Brittany collections ID (P4313) be I have to work backward afterward so it will take a bit of time, it should be fixe soon). When it comes from the portal, should I repeat it for every statements?
country (P17) is absolutely necessery, we already talked about that ;) in France, (most) museum objects are under directly the sovereignty to the state (and not the museum itself), it is an important data.
@VIGNERON en résidence: I think that most art historians are not familiar with the work of Joseph Oberthur and Simone Le Moigne - even their names are probably unfamiliar to most scholarly trained art historians. Of course, this doesn't mean that they are bad painters, but the qualification "rather obscure" is quite fitting, globally speaking. You could say the same about many painters in the Musée d'art moderne et contemporain de Strasbourg and the Musée des Beaux-Arts de Mulhouse, it's nothing personal or against Britanny! --Edelseider (talk) 09:50, 10 January 2023 (UTC)
@VIGNERON en résidence: I don't like the "stored in the Brittany Museum" in the description of every painting. It's superfluous and it feels a bit like you're trying to promote the museum. It sure is completely not in line with the 600.000+ other painting items we have here on Wikidata. Can you update the description? Multichill (talk) 18:36, 9 February 2023 (UTC)
@Multichill: Done, is it ok now? (I also tried to add a description in German, Spanish and Italian). Is there any hint anywhere of what is a good "description"? (ideally in multiple languages) PS: I see that a lot of other paintings with the museum in description (qv. [3] ; and it's seems to be the usual practice for non-painting objects). Cheers, VIGNERON en résidence (talk) 09:23, 10 February 2023 (UTC)
Italy is currently listed as having less than 5,000 paintings on its entire territory. Who among us doesn't know that in Florence alone, you find at least three times that number, spread over the museums, the churches, the chapels, and the palaces? Sailko, il tuo lavoro è necessario qui. L'Italia è sottorappresentata! --Edelseider (talk) 14:32, 9 January 2023 (UTC)
I provide pictures.. providing also Wikidata entries is too much for me ;) But I recently sorted my images by work of art represented, here, it helped some users to improve templates and wikidata stuff. --Sailko (talk) 15:22, 9 January 2023 (UTC)
I was looking at the statistics and looks like User:Stevenliuyi imported 50.000+ paintings from the The Palace Museum (Q2047427). Happy to see a large import of paintings that are not in the European tradition! No trial run was done, but I think some fixes need to be applied:
Hi, I'm still working on it, especially uploading those with public domain images to Commons and linking to the items here using image (P18). There are also data for "level of cultural relics", which I guess could be imported using heritage designation (P1435). Regarding the above-mentioned fixes, I'll gladly make the necessary changes once a consensus has been reached on how to model Chinese paintings, which may need some more consideration to conform to the model primarily used for Western paintings. For instance, traditional Chinese paintings can be divided into two board categories, ink wash painting (Q756712) and ink and color painting (Q116457653). Perhaps using made from material (P186) instead of genre (P136) would be better, but I'm not sure the best way to incorporate the information into made from material (P186).--Stevenliuyi (talk) 19:57, 9 February 2023 (UTC)
To the points raisded by Multichill in the original inquiry:
In our Wikidata:WikiProject_PCC_Wikidata_Pilot/Smithsonian_Libraries/Projects/Chinese_Portraits (CAPP), we followed a similar trend of practices with some variations. Our guidance relies on recording the statements based on published authoritative database for East Asian, namely Chinese, paintings and existing Wikidata practices as we came across throughout the Wikidata:WikiProject_PCC_Wikidata_Pilot period (Sep 2020 to Aug 2021). In our experience, many descriptions applied to Western paintings were either insufficient or somewhat inappropriate for East Asian works, especially the Chinese portrait paintings. Though, we have listed selective properties for our paintings project, we have not always succeeded in applying consistently across the project as they were not core.
The Chinese ancestor portrait is a distinct style of painting. Though such Q item "ancestor portrait" is absent, another item describing the type of painting is available, portrait (Q134307) which our team opted to use for genre (P136). In hindsight, we could have created an item for the "ancestor portrait"
At times, the Chinese style of paintings, though some interprets as <genre>, also encompasses painting technique. For example, shui-mo or shuimo (ink wash or ink and wash), in early periods usually only ink (black). There are also Portrait or Figure painting, Landscape painting and Bird-and-Flower painting. The other things we know about are “styles” or “schools of painting; ie. "Literati painting," "Throw Ink," or “Written Idea” freehand style be called shuimo, xieyi) see WP article.
We have used the property country of origin (P495) to express period of the dynasty when a specific date cannot be ascertained, ie, Portrait of Daisan (Q107650956). Property of time period (P2348) was opted for items with no definitive reference in any of the published authoritative sources. Since the collection curators have already identified the group paintings were from Qing period, we use that property, time period (P2348), to narrow the timeline in order to facilitate query output. Another property we have chosen to incorporate, owned by (P127) (showing once upon a time ownership), ie, A Breath of Spring (Q18689538). This allowed us to bring together the collection we hope to highlight.
We have also chosen the property made from material (P186) to describe the characteristics of painting materials that the artist used
As for theme of the painting, we have chosen the property depicts (P180) to convey what the painting was about, the sitter, flowers, birds, rosary, etc.
Did anyone notice that, for a few days, LB went into overdrive and updated everything every 24 hours or fewer? Well, now it even fails to log in. If you request a manual update, you get the message: "ERROR: Login failed". This tool drives me crazy! Edelseider (talk) 19:24, 9 February 2023 (UTC)
Edelseider I do rely on ListeriaBot a lot. For fast and dirty list it is great; however it does have many things it does not do correctly, like for example displaying or sorting based on nontrivial dates. One alternative I use is to write custom lua code to display each work (I cobbled together c:Module:User:Jarekt/table row to display information in Polish), and use SPARQL query to get a sorted list of artworks to show. Then I use a spreadsheet to add display template (with or without "subst:") to each artwork Q-code. It is a bit more manual work and not as convenient, but I do have full control over the results and can create tables ready to be added to articles, like in pl:Lista dzieł malarskich Stanisława Ignacego Witkiewicza. --Jarekt (talk) 12:30, 13 February 2023 (UTC)
Note you can always pick out the query in the "Select" statement from your listeria list and dump that Sparql query into the Wikidata query service for updating your list of items. The query service doesn't include many of the features of listeria, but it at least provides basic feedback if an item is or is not in a list. I use that when listeria goes down, in combination with the good old "What links here" button (Special:WhatLinksHere). Jane023 (talk) 10:46, 15 February 2023 (UTC)
For the record, we have a lot of loan here at the Museum of Brittany (Q3329701), I see that we already have ~300 items on Wikidata about objects where I could add this data (in France it's quite easy to find them as the data is already implicitely stored in the inventory number (P217), inventory number for loans has to start with a D).
Hello VIGNERON en résidence. Indeed it's often easy for French museums to guess with the number that an artwork is on long-term loan. Unfortunatly the collection which holds it is not always mentionned. For wikidata edits, imho, the model of
is a good one. With more than 4000 claims, even if it could be considered as quite low, it's already a practice of edition. Best regards --Shonagon (talk) 23:50, 8 February 2023 (UTC)
@VIGNERON en résidence, Shonagon: loans modeling needs some attention. We have the concept of a "loan" and the concept of a "long-term loan", but the distinction is currently unclear.
@VIGNERON en résidence @Multichill @Jane023 The distinction between "loan" and "long-term loan" is very clear for museum professionals for whom these are common practices. In French there is 2 two differents words for these 2 different things: "prêt" for "loan", and "dépôt" for "long-term loan". One of the big differences for long-term loan is that the loaned object becomes a part of the borrower’s collection for the time of the loan. That's why we have to use in Wikidata the collection (P195) for both collections, lender and borrower, and to qualify the long-term loan. The responsibility for preservation is shared with the lender, even if the work is on long-term loan for decades. For a example if a painting is on "long-term loan" in an another museum or in a church (common cases), the museum to whose collection the work belongs, ensures regularly that the work remains in good condition, continue to document the work and manages legal provisions such as inventory registration.
@Shonagon: your response seems to be very specific to France and French museum professionals. commodate (Q1580844) / nl:Bruikleen is the term used in the Netherlands.
Distintion between loan and long-term loans is not specific to France. The Long-term Loans and Collection Research Working Group 2008–2010, a subgroup of the European Open Method of Coordination Working Group on Mobility of Collections made a Long-term loans best practices report (pdf) in which there is clear distinction between short loan and long-term loans: Long-term loan practices differ from those of short-term loans in many respects. The most noteworthy difference is that objects on long-term loan are meant to contribute to the borrower’s permanent collection and are regarded as forming part of the collection for the period of the loan. Objects on long-term loan are thus treated with the same care, use and attention as all objects in the borrower’s collection. A more precise description of the differences between long-term and short-term loans is attached to this report as Appendix A . There was no French people in this working group, but museum professionals from Austria, Belgium, Bulgaria, Germany, Italy, Luxembourg, Malta, Portugal, Spain and Romania.
Would me nice to get this sorted out. Plenty of loans that are currently not documented as being a loan. @Jane023: didn't we have a conversation about this before? Multichill (talk) 19:00, 25 February 2023 (UTC)
hm yes, I forgot about this. I guess it would be nice to have the same method for short or long term loans. In the Netherlands some long-term loans are a few years but others can be decades. I recall wanting an item for a loan object - should these be long-term loan as soon as the loan is longer than for an exhibition? Jane023 (talk) 19:33, 25 February 2023 (UTC)
Sorry forgot again about this. Yes I noticed that there were a few different properties for loans of sportsplayers. Some art loans outlast the expected human life term though. It was after poking around those contractual agreements that it occurred to me we should do something about those. Very common (see Vermeer case ongoing) is short-term loan exchange where certain works get traded for others (think Brad Pitt in Moneyball juggling players around the league to control his team manager). It would be nice to have "art loan" the way we also have "art theft" but instead of a significant event, some sort of collection-related qualifier. Jane023 (talk) 07:26, 17 March 2023 (UTC)
Done - note these are not items, but pages, so you can just make a redirect like any Wikipedia page (not sure if you are active on Wikipedia or not, but for future reference for anyone else). For common names of collections in and out of the host country, it's probably worthwhile making sure that that if there's ãt least one list article, a redirect in either the host country language or English would be useful to have in the Wikidata lists. Jane023 (talk) 10:41, 15 February 2023 (UTC)
As of today, 799 of Pablo Picasso's paintings have been given the WD treatment. That's about one third, or 33%, of his total output as a painter. Shouldn't it be an absolute priority that the list of his works be completed? He shouldn't be ranked so low here: Wikidata:WikiProject sum of all paintings/Top creators by number of paintings.
Speaking of notable painters who have left us tons of paintings without that fact being reflected here, let us pause one moment and think of poor Jean-Jacques Henner. That man has covered well over 1,000 canvases and panels in paint and even has a National Museum to his name. And yet, he doesn't even appear on that list! Life is unfair. Edelseider (talk) 09:15, 29 March 2023 (UTC)
@Edelseider: It's also about volunteers! Each online or offline resource needs a Wikidata volunteer to work on importing the metadata locked inside. For paintings by local painters I am always interested. For paintings by painters outside my local or favorites list, not so much. Jane023 (talk) 08:09, 27 May 2023 (UTC)
Syndication of other online catalogs
We currently link to quite a few other online catalogs. I've been contemplating for a while if we should do more bulk imports of these kind of catalogs. These items would probably be of lower quality (less statements) than what I normally do. It would increase the number of paintings and thus the coverage we have making it easier for others because than you have an existing item to expand. Any thoughts? Multichill (talk) 10:27, 7 May 2023 (UTC)
SOAP relies on bots, and these bots are working less and less well. It's frustrating.
@Jane023: @Multichill: @Jean-Frédéric: @Nono314: @Thierry Caro: We've had one of these weekends again where nothing works. InteGraalityBot has overslept ([4]), and BotMultichillT has, once again, not proceeded to the weekly update ([5]). Add to this that ListeriaBot doesn't do any manual updates any more. We are too reliant on bots, and now we are f*cked. Edelseider (talk) 07:01, 2 October 2023 (UTC)
My listeria lists are all working semi-regularly, so if yours haven't run in a while, try trimming them down - they may be timing out. I regularly trim my "Top <whatever>" lists that use counts, by increasing the count by one. When I get annoyed I just grab the query out of the listeria list and run the query directly and go from there, which is often faster anyway. Jane023 (talk) 07:12, 2 October 2023 (UTC)
If it was only Listeria - but the others bots are working less well, too. That's the point of my complaint... three bots are malfunctioning at the same time. And I didn't even check HannolansBot. Edelseider (talk) 07:50, 2 October 2023 (UTC)
Thanks for bringing this to my attention. integraality currently tracks close to 800 dashboards ; 665 were correctly updated last night according to the execution logs.
(As a tangent − while I do understand the sentiment, comments like “We are too reliant on bots”, “We are f*cked” etc. have somewhat mixed effects on my motivation to work on integraality… Jean-Fred (talk) 14:53, 2 October 2023 (UTC))
Done, is it ok now? (I also tried to add a description in German, Spanish and Italian). Is there any hint anywhere of what is a good "description"? (ideally in multiple languages) PS: I see that a lot of other paintings with the museum in description (qv. [3] ; and it's seems to be the usual practice for non-painting objects). Cheers, VIGNERON en résidence (talk) 09:23, 10 February 2023 (UTC)
⟨ said collection ⟩
