[04:19:49] If this thought is not wrong, then perhaps it is misguided? https://twitter.com/stuartayeates/status/1433162135622889479 [05:34:02] looking at the replies, are they both under the impression that we still use p107 or something? [05:48:58] and the linked blog post, that was 7 years ago, which was so long ago that adding statements had only been possible for a year... [07:48:14] Writing it is "an attempt" sounds as if that is an official goal or so. [07:48:14] But the concern itself is imho very valid – "Universal Abstract Knowledge" is a very western idea afaic, tied to very specific ideas why abstract is desireable, and on how thinking, language and writing work. [07:48:16] The concerns about this in anthropology and history of knowledge date back at least 20 years e.g. Adam's "Artificial Knowing", Bowker’s "Keeping Knowledge Local" and Turnbull’s "Working with Incommensurable Knowledge Traditions". (re @mahir256: If this thought is not wrong, then perhaps it is misguided? https://twitter.com/stuartayeates/status/1433162135622889479) [08:00:36] Is the notion being largely Western in nature inextricable from any flaws in previous attempts attributable to a particularly Western idea insurmountable within the designs of those attempts? [08:06:29] Sorry, I do not understand the question – could you take it apart a bit? [08:10:53] You say that failures of previous attempts might be attributable to western concepts; the question is if such flaws might be ruled out or not (then they are insurmountable) – correct? [08:15:31] is it cause you're Western? :-D /s (re @Jandit: Sorry, I do not understand the question – could you take it apart a bit?) [08:18:30] These previous attempts that were created by Westerners likely had some flaws that originated due to a particular conception or assumption, that was specifically Western in origin, which could not be surmounted; is the existence of such a conception/assumption guaranteed for any future system developed in part by Westerners? [08:19:35] (or, alternately, is Denny being Croatian in ancestry and German in education a sign that some distinctly European phenomenon in Abstract Wikipedia's development will cause it to fail?) [08:23:44] Ah, thanks for helping me here. [08:24:15] should the answer not be a categorical "yes", that suggests that within the development of this new system there may well be ways to overcome such potential challenges beyond the abandonment of the effort to which the Twitterer seems to allude, and that our goal should be to stay vigilant regarding these challenges [08:25:50] I do not think that the involvement of any particular individual (or their educational background) causes the fail of a system. [08:27:56] The above authors criticizing "databased abstract universal knowledge" do not refer to any individuals and their particular backgrounds either. [08:28:41] Their framing of the problems is focussed on infrastructure and history. [08:32:33] so when the Twitterer speaks of a "Western ontology" being "imposed", the term "Western" has nothing to do with ideas which were predominant on the continent (and thus propagated by its inhabitants)? (re @Jandit: Their framing of the problems is focussed on infrastructure and history.) [09:18:59] can't get list equallity to behave when the list contain code points? Any idea why? [09:29:48] https://phabricator.wikimedia.org/T254695 seems upstream now..any docs for the loading indicator library? [09:54:34] !admin@notwikilambda is notwikilambda orchistrator down? [09:54:34] Sorry, you are not authorized to perform this [09:56:24] :( [09:57:10] !admin@wikipedia-abstact ^^ [10:16:11] I can take a look at the orchestrator later [10:24:18] ok! [10:25:05] now it is up though :P [15:29:36] oh, I guess you might have tried to use the orchestrator while it was busy restarting itself to pick up potential updates [15:30:19] in a production setup this would happen without downtime, of course, but this isn’t a production setup :) [15:30:30] maybe I’ll look into it as a kubernetes exercise [15:47:52] "is an builtin equality for..." <- I want to help with this.. [19:14:12] The newsletter this week is written by Mahir256 - thanks! On generating text with Ninai and Udiron [19:14:12] https://meta.wikimedia.org/wiki/Abstract_Wikipedia/Updates/2021-09-03 [19:46:30] @jimman2003:matrix.org - re: is an builtin equality for integers wanted? - I would not want it as a builtin. I don't even want the integer type as a builtins, but want that to be community defined - which means integer equality would also be community defined. [19:48:10] re @jimman2003: seems like Z10073 works (for strings and bool at least).. - Yay! It looks good :) Thanks! [19:51:01] not for code points though (Z10073) i mean [19:53:22] /me wonders if we want Z-ID support in wikilinksbot [19:53:33] I suppose it’s a bit early while the real wiki hasn’t launched yet [19:53:51] true [19:54:50] (also, presumably it wouldn’t work well across the IRC bridge, because of this Telegram thing where bots aren’t able to see other bots’ messages) [19:56:52] s/not for code points though (Z10073) i mean/not for code points though, the latter/ [19:57:47] @jimman2003:matrix.org - re: can't get list equallity to behave when the list contain code points? Any idea why? -- I think that's due to https://phabricator.wikimedia.org/T290122 - i.e. Code Points don't get translated to Python yet [19:58:28] oh,didn't connect the dots [20:14:18] That would be nice! It would still work on the telegram side (or IRC side), depending where your not runs (re @lucaswerkmeister: /me wonders if we want Z-ID support in wikilinksbot) [20:16:28] I think, given that argument, the issue with "Universal Abstract Knowledge" is not so much in the "Abstract" but already in the "Universal Knowledge" part. (re @Jandit: Writing it is "an attempt" sounds as if that is an official goal or so. [20:16:29] But the concern itself is imho very valid – "Universal Abstract Knowledge" is a very western idea afaic, tied to very specific ideas why abstract is desireable, and on how thinking, language and writing work. [20:16:31] The concerns about this in anthropology and history of knowledge date back at least 20 years e.g. Adam's "Artificial Knowing", Bowker’s "Keeping Knowledge Local" and Turnbull’s "Working with Incommensurable Knowledge Traditions".) [20:17:46] The idea of an encyclopedia in the first place is an idea of enlightenment, which originates in the history of Europe. [20:19:24] I doubt that there will be arguments that are particular to Abstract Wikipedia, which wouldn't have been valid to Wikipedia in the first place. [20:21:26] And if the argument is "we shouldn't have a Wikipedia in Maōri in the first place, and therefore we shouldn't have Abstract Wikipedia", then the argument is consistent, but I disagree with it and think it's valuable to offer more access to more knowledge that anyone can edit, in their own language. [20:24:54] trying my hand at T290122... [20:28:08] can sb give me a code point to test my code with? [20:31:57] Most will apply to both, this is correct. Adams "Artificial Knowing" has a pretty good critique of cyc that imho also applies to Wikidata. (re @vrandecic: I doubt that there will be arguments that are particular to Abstract Wikipedia, which wouldn't have been valid to Wikipedia in the first place.) [20:35:23] Ab? (re @wmtelegram_bot: can sb give me a code point to test my code with?) [20:35:32] sb? (re @wmtelegram_bot: can sb give me a code point to test my code with?) [20:35:39] somebody [20:35:50] s/sb/somebody/ [20:36:02] s/can sb give me a code point to test my code with?/can somebody give me a code point zobject to test my code with?/ [20:37:02] Hmm, CreateZObject on Notwikilambda doesn't like me right now, otherwise I'd create it [20:39:01] yeah,broken atm [20:39:03] Here's how a codepoint looks like in JSON, if that's helpful: { "Z1K1": "Z86", "Z86K1": "a" } [20:39:28] * yeah,broken atm [20:39:34] * jimman2003 uploaded an image: (35KiB) < https://libera.ems.host/_matrix/media/r0/download/matrix.org/tQHUvSHWdBcjWUyXCuMkHqzo/image.png > [20:39:44] vrandecic: ty [20:40:18] * ty , that was what I was asking [20:42:15] so this should work ```_DESERIALIZE_Z86 = lambda Z86: Z86['Z86K1']... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/701bd07c4fa32d97e10a64437fa96da4171c61f1) [20:42:20] * so this should work ```_DESERIALIZE_Z86 = lambda Z86: Z86['Z86K1']... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/ea57a463cd26e95001dae294c80e8a26f69edb8f) [20:42:42] * so this should work... (full message at https://libera.ems.host/_matrix/media/r0/download/libera.chat/8155d289ae9dc08cb60e10b9faa49458519e87b9) [20:47:52] Maybe :) That's a part of the code I am not very familiar with, and Cory, who would know best is on vacation. Maybe @James_F [20:52:51] well,https://gerrit.wikimedia.org/r/c/mediawiki/services/function-evaluator/+/717596 [20:53:02] s/,/ / [22:39:57] Thanks! [23:59:18] CreateZObject is fixed on Notwikilambda (Thanks Lindsay for the patch!)