inventaire.io @inventaire@mamot.fr

@alex
1/ oui, en plaçant le curseur dans la barre de recherche, ça devrait marcher
2/ ce serait effectivement absurde, et c'est pour ça que je pense que les données bibliographiques doivent rester centralisées (voir https://github.com/inventaire/inventaire/issues/187 )
3/ actuellement non, mais de meilleurs moyens de catégorisation sont une des demandes les plus récurrente, on y travail donc mais ça va prendre du temps pour faire ça bien
4/ le seul format d'export géré aujourd'hui est le JSON https://inventaire.io/settings/labs

@alex la documentation ne doit pas être interprété comme une invitation à l'auto-hébergement d'une instance, le soft n'est pas prêt/optimisé pour ça, il faudrait rendre la documentation plus clair sur ce point -- Maxime

@aldolinux @framasky Inventaire permet de créer ces propres œuvres si l'ISBN n'est pas reconnu par le scan : soit en dessous d'une recherche infructueuse dans la barre principale, soit en cliquant sur "ajouter une œuvre" à partir d'une page d'auteur.

@Readlebee all descriptions come from Wikidata (Inventaire entities don't have descriptions fields). There have been discussions within the #Wikidata community on auto-generating descriptions, but nothing is settled afaik. I had been doing some scripting to mass add book descriptions though https://github.com/maxlath/wikidata-scripting/tree/master/books_descriptions -- Maxime

@Readlebee yes! example: https://inventaire.io/entity/inv:153fc32a4ed14f65e3bf5afd9b80b51e
no worries, send your questions :) #inventaireio

@ahstro @Readlebee hard to answer that question, given that we actually use both, and that both have different qualities. but has Inventaire was started by linked open data enthusiasts, Wikidata was the way to go :D some things would have been simpler if we went with OL (mostly book covers + that an entity can't change type randomly as we see with WD :p) but working with Wikidata also brings a lot of goodies ^^ (links to Wikipedia, author graphs, characters lists, etc)

@Readlebee the more people use the data, the more they are interested in helping to improve it, so we are generally very open to other libre projects making good use of our resources :)

@Readlebee the front end live search uses this endpoint https://inventaire.io/api/search?types=works&search=Transmetropolitan&lang=en&limit=10 which is a direct call to an ElasticSearch instance where works are indexed with their editions book covers and the re-sorted to take popularity into account https://github.com/inventaire/inventaire/blob/90ef593/server/controllers/search/search.coffee#L26-L29

@tootbrute @Alamantus yep, we have the same frustration here, things going too slowly on the interface; our excuse is that we have been working on ways to improve the data (through admin interfaces to mass merge entities for instance). Also, those days, we are busy working for the French National Library, so we have less time for inventaire, but that ends in November. -- Maxime

@Alamantus @dmoonfire we follow Mastodon address format as expected by #Wikidata https://www.wikidata.org/wiki/Property:P4033#P1855 so that would be dmoonfire@octodon.social
the interface should probably help figuring it out

@Alamantus so for instance, if you want to find the book covers associated with the work https://inventaire.io/entity/wd:Q47598 you can request all its editions uris https://inventaire.io/api/entities?action=reverse-claims&property=wdt:P629&uri=wd:Q47598 and then all the data from those uris https://inventaire.io/api/entities?action=by-uris&uris=isbn:9780439655484|isbn:9782070528189|isbn:9781856136174|isbn:9780613359580|isbn:9780439612555|isbn:9781408845660|isbn:9781551922461|isbn:9788183220309|isbn:9781408855676|isbn:9781408856772|isbn:9781338299168|isbn:9781856136631|isbn:9780747560777|isbn:9789667047368|isbn:9782070621989|isbn:9780439136365|isbn:9780545162074|isbn:9780545791342|isbn:9785353004325|isbn:9782070584925|isbn:9783551551696|isbn:9781594130021|isbn:9781435238121|isbn:9780747546290|isbn:9782070612383|isbn:9782070643042|isbn:9782070529292|isbn:9780545596275|isbn:9780747595847|isbn:9780747542155|isbn:9782075089302|isbn:9781408812839|isbn:9788372786982|isbn:9781781100233|isbn:9780747573623|isbn:9789513117375|isbn:9782070541300 with which you should be able to filter by language (entity.originalLang) and find the book covers ids (entity.claims.invp:P2[0]) from which you can build an image URL https://inventaire.io/img/entities/11d49f86b9782a6790022c7a87699661862bbfcd #inventaireio #api #wikidata

@Alamantus the important concept there is the distinction between works and editions: works are the abstract idea of a book, it doesn't have a book cover, while editions are the instantiation of that work by a publisher and thus has a book cover (see entities map https://inventaire.github.io/entities-map/)

@Alamantus @libreture here are also JS libs we developed to work with #wikidata :
- read/parse https://github.com/maxlath/wikibase-sdk
- edit https://github.com/maxlath/wikibase-edit
- CLI https://github.com/maxlath/wikibase-cli

@Alamantus @libreture some reading recommendations ;)
- the project where people coordinate the modelization of books data in #wikidata https://www.wikidata.org/wiki/Wikidata:WikiProject_Books ,
- see also #Wikicite http://wikicite.org
- how #inventaireio uses and extends those data https://wiki.inventaire.io/wiki/Data

Statuses of the association has been translated in english https://wiki.inventaire.io/wiki/Association?lang=en As a first draft, every contributions for english correctness is much welcomed

@Alamantus @switchingsocial@mastodon.at @libreture lists of books (either wishlist, reocmmandations, bibliographies, etc) are something we are working on to integrate to Inventaire: I'm quite certain it will happen, it's just hard to say when. If you're interested in helping us building it, you're all very welcome to join our communication channels https://wiki.inventaire.io/wiki/Communication-channels -- Maxime