[18:25:55] A nice, tangible outcome from Wikimania Hackathon was a poster advocating for more color and textured 3D formats for Commons, and the discussion saw widespread support for it. [18:25:56] https://commons.wikimedia.org/wiki/Commons:Village_pump/Proposals#Enable_textured_3D_files_on_Commons [18:25:58] [18:25:59] Not sure what the next move is but here is the Phabricator: [18:26:01] https://phabricator.wikimedia.org/T246901 [18:29:37] Unfortunately it's one of a stack of basically abandonware WMF written/deployed MW extension [18:46:42] In theory, the upgrading of Thumbor means one of the blockers is out of the way. But I'm not sure about the rest (re @tehreedy: Unfortunately it's one of a stack of basically abandonware WMF written/deployed MW extension) [19:02:28] https://www.mediawiki.org/wiki/Manual:Adding_support_for_new_filetypes [19:05:13] These are the general steps. The biggest problem is that any of this is just a lot of work. Upgrading thumbor took an entire team almost a year. Releasing the new video player (which mostly had already been written) took 6 months of ‘launch work’ by 6+ people (refinement, community communication and refinement) etc [19:07:12] You need to implement upload and format parsing, update the ‘player’ with new libs that support these formats, security review, thumbnailing support etc. I estimate at least a year of work, unless u find someone very motivated versed in MediaWiki/WMF [19:12:41] BTW, this is also a nice time to remind folks that there is no pointer to "3D" content from the front page of Commons and because it's not considered "media," navigating to 3D content is not obvious or easy. [19:14:11] 3dtopng is even more abandonware than thumbor was [19:14:27] Absolutely [19:17:37] I suppose a good first step could be to implement geometry-only parsing for a format, other than .obj, which supports materials/colors/textures — once the geometry is loaded into memory, the rest of the code for handling obj files (displaying, manipulating, thumbnailing, etc.) should be reusable. Then a second step would be to actually support non-geometry aspects of the new fo [19:17:37] rmat. But I'm sure there I'm missing complexity that means it wouldn't be this straightforward 😅 [19:20:33] Oof, is Thumbor considered abandonware? The github repo doesn't show tons of activity. So basically Thumbor was upgraded from ancient to just old? (re @djhartman: Absolutely) [19:20:48] there is an upstream again, there wasn't for a while [19:21:22] but the single WMF person maintaining our custom engines left, and I refused to maintain it as a volunteer on principle. [19:26:59] For all the words by WMF management, the support for basic Commons functionality is still pretty much as shit as it was 5 years ago. Its just piecemeal while it should be dedicated and continuous. [19:31:24] And they deleted [[File:Asd.gif]] 😭 (re @djhartman: For all the words by WMF management, the support for basic Commons functionality is still pretty much as shit as it was 5 years ...) [19:36:00] (That - for the records - it was this very important file, in use in every documentation page of Wikimedia Italy - at least http://web.archive.org/web/20230408091339id_/https://upload.wikimedia.org/wikipedia/commons/1/13/Asd.gif ) [19:39:36] Because of the extensive CI that the 3D extension has, https://gerrit.wikimedia.org/r/c/mediawiki/extensions/3D/+/955983 actually passes, but I really doubt it would [19:51:37] I checked, WMF did not. The Commons community did, because it was a copyright violation. Please keep your facts straight (re @bozzy: And they deleted [[File:Asd.gif]] 😭) [19:51:52] (I was jocking) (re @djhartman: I checked, WMF did not. The Commons community did, because it was a copyright violation. Please keep your facts straight) [21:48:51] IMO it's just not very likely that 3D models would produce significant mission impact (bring many new readers, or new editors, or significantly increase the content's usefulness for readers) so they aren't very competitive with other things one could spend money on. (re @djhartman: For all the words by WMF management, the support for basic Commons functionality is still pretty mu [21:48:52] ch as shit as it was 5 years ...) [21:50:13] The same is true for many Commons things. Not basic upload, the usefulness of that is clear, but e.g. what's the practical difference between the image behind a Wikipedia thumbnail being 100MB or 10GB (other than the second one being more costly for us)? [22:56:42] Depends on whether we keep the conservative and narrow mission of today, or whether we really believe the high-minded rhetoric we have in the Strategy - "By 2030, Wikimedia will become the essential infrastructure of the ecosystem of free knowledge, and anyone who shares our vision will be able to join us." [22:56:43] [22:56:44] I would say supporting widely used, not terribly exotic 3D files is reasonably part of the "ecosystem of free knowledge" (re @gtisza: IMO it's just not very likely that 3D models would produce significant mission impact (bring many new readers, or new editors, o...)