[02:12:57] I pronounce it fiːɡələɡːif/FEE-guhluhg-eef (not that it's offensive or anything-it's a pseudonym-I just think fiːɡələɡːif in my head, so it'll make it easier to communicate) (re @Sannita: The recording of Monday's Volunteers' Corner is now available on Commons: https://commons.wikimedia.org/wiki/File:Abstract_Wikip...) [02:31:08] And one more thing, @vrandecic, what drafted type proposals are you looking at creating next? [05:53:18] I'd prefer staff use volunteer accounts for everything that doesn't have a phabricator task and is not an office action, but I doubt that's the position that's shared by many (I think the WMF has done pretty huge overstepping, e.g. superprotecting on dewiki, banning Fram, following the orders of the court in ANI vs. WMF, trying to call themselves the "Wikipedia [05:53:18] Foundation", their [05:53:18] fundraising banners that suggest that the WMF is financially in trouble when they are perfectly afloat, spending said donation money on grants instead of using it for wiki projects, is actively far-left leaning (and I say that as someone on the left), the list goes on ) (re @vrandecic: We're happy to follow the policy the community comes up with. I don't think the [05:53:18] current situat [05:53:19] ion is well-thought through.) [06:30:54] I have a very different take. Volunteer accounts should only be used for edits not done as employees. Anything done as staff should use the staff accounts. (re @Feeglgeef: I'd prefer staff use volunteer accounts for everything that doesn't have a phabricator task, is not an office action, and is not...) [06:32:05] Yes. I agree with you. Staff should be doing less on-wiki stuff as staff (re @Jan_ainali: I have a very different take. Volunteer accounts should only be used for edits not done as employees. Anything done as staff sho...) [06:32:58] Something like "per phab:T123456" is reasonable, something like "add en label" on a random function is not. [06:33:08] /delete@wikilinksbot [06:33:32] You just said that staff should use volunteer accounts? (re @Feeglgeef: Yes. I agree with you. Staff should be doing less on-wiki stuff as staff) [06:35:08] I said that something with a Phabricator task or that's like a status update is fine (re @Jan_ainali: You just said that staff should use volunteer accounts?) [06:36:01] Where I disagree is, again "add en label." In my opinion, that should be exclusively on volunteer accounts. [06:36:03] If you meant "people who also are staff should use volunteer accounts outside work hours", then yes, we agree. [06:36:49] And, they shouldn't make mainspace changes unless they are part of official and well-documented work [06:38:18] Like, working on a status update, or making a type you said would be created, and fixing an issue on Wikifunctions is fine. I don't think staff should be making any other edits inside work hours. [06:43:50] Oh, from your first message it sounded like you _expected_ staff to do edits during work hours but with their volunteer accounts. (re @Feeglgeef: Like, working on a status update, or making a type you said would be created, and fixing an issue on Wikifunctions is fine. I do...) [06:44:57] I'm working on float64 (re @Feeglgeef: And one more thing, @vrandecic, what drafted type proposals are you looking at creating next?) [06:46:28] Can you connect Z20780? (re @Feeglgeef: @vrandecic can Z20780 be connected as a renderer for Z20420? [06:46:28] I've made some functions for it: [06:46:30] Z20770 for Dagbani [06:46:31] Z20773 for Engl...) [06:47:16] I see two conditions before giving out maintainer rights: 1) having a more robust community, 2) having an updated type editing experience (re @Feeglgeef: @Sannita @vrandecic can I get an answer to this?) [06:48:43] good, I'm wondering when that would be tho... (re @vrandecic: I see two conditions before giving out maintainer rights: 1) having a more robust community, 2) having an updated type editing e...) [06:50:24] I will give you that our community is not very robust. 2 people have edited the mainspace in the last 24 hours; me and you. (re @vrandecic: I see two conditions before giving out maintainer rights: 1) having a more robust community, 2) having an updated type editing e...) [06:50:25] I would say the first condition is firmer than the second [06:51:21] I would expect that the community will continue grow as we get integrated into Wikipedias and as the system itself becomes more robust and functional [06:52:42] done. Let's see how that works 🙂 (re @Feeglgeef: Can you connect Z20780?) [06:52:55] Thanks! (re @vrandecic: done. Let's see how that works 🙂) [06:54:20] Yes; I'm betting that communities will notice things that we are missing and try to add some. (re @vrandecic: I would expect that the community will continue grow as we get integrated into Wikipedias and as the system itself becomes more ...) [07:09:02] Q104180541 [07:20:11] speaking of type, what type would be the inputs for a conjugation function? (eg. if I want to compose all the small conjugation function of https://www.wikifunctions.org/wiki/Wikifunctions:Catalogue/Natural_language_operations/Breton into one general function) [07:23:59] Mainly string or lexeme [07:24:09] Natural number in some cases [07:25:33] Boolean in some others [07:26:03] string seems bad and inefficient (and not multilingual, skrivet e vo e Brezhoneg neuze :/ ) (re @Feeglgeef: Mainly string or lexeme) [07:26:09] lexeme? what/why/where? [07:26:26] which cases? for the grammatical numbers? (re @Feeglgeef: Natural number in some cases) [07:26:49] Yes, or Subjects (re @Nicolas: which cases? for the grammatical numbers?) [07:27:34] what is "Subjects"? (re @Feeglgeef: Yes, or Subjects) [07:28:21] 1st person (I), 2nd person (you), 3rd person (Them) (re @Nicolas: string seems bad and inefficient (and not multilingual, skrivet e vo e Brezhoneg neuze :/ )) [07:30:29] ah, I've never heard that called "Subjects" before, but okay, in Breton it's 0th person, 1st person singular, 2nd person singular, 3rd person singular masculine, , 3rd person singular feminine, 1st person plural, 2nd person plural, 3rd person plural masculine, 3rd person plural feminine (re @Feeglgeef: 1st person (I), 2nd person (you), 3rd person (Them)) [07:31:12] You've never heard of the subject of an action before? (re @Nicolas: ah, I've never heard that called "Subjects" before, but okay, in Breton it's 0th person, 1st person singular, 2nd person singula...) [07:31:50] plural is a Boolean and feminine is a Boolean (re @Nicolas: ah, I've never heard that called "Subjects" before, but okay, in Breton it's 0th person, 1st person singular, 2nd person singula...) [07:31:55] yes, but the grammatical numbers is not directly based on the subjects, that's weird to mix them (re @Feeglgeef: You've never heard of the subject of an action before?) [07:32:56] which begs the question, it one or three inputs : the list above or number/plural-singular/masc-fem [07:32:57] the second could lead to strange things (like "0th person singular masculine" which doesn't exist) (re @Feeglgeef: plural is a Boolean and feminine is a Boolean) [07:35:42] 0th person singular masculine is we ignore the boolean (re @Nicolas: which begs the question, it one or three inputs : the list above or number/plural-singular/masc-fem [07:35:43] the second could lead to st...) [07:36:15] We *could* throw here if we don't get false-false but I don't think that's the best move. [07:37:57] obviously, but is it a good idea to have input that will be ignored most of the time? (re @Feeglgeef: 0th person singular masculine is we ignore the boolean) [07:38:51] I'd imagine it doesn't matter that much to prefill the values (re @Nicolas: obviously, but is it a good idea to have input that will be ignored most of the time?) [07:40:27] it depends who is using the function... (re @Feeglgeef: I'd imagine it doesn't matter that much to prefill the values) [07:41:38] This is the kind of thing that compositions can't replace. While it does remove the syntax bit, creating Wikifunctions compositions still requires this kinda problem solving given a scenario, which I think will make it hard for people with little technical knowledge to use well. (re @Feeglgeef: We *could* throw here if we don't get false-false but I don't think [07:41:38] that's the best move.) [07:41:53] Who do you see using it/in what contexts? (re @Nicolas: it depends who is using the function...) [07:43:54] honestly, no-one (Breton is an almost dead language and we mostly don't use conjugation) [07:43:55] but I'm thinking that it could be anyone (re @Feeglgeef: Who do you see using it/in what contexts?) [07:44:59] Nationalism at its worst :( (re @Nicolas: honestly, no-one (Breton is an almost dead language and we mostly don't use conjugation) [07:45:00] but I'm thinking that it could be anyon...) [07:46:04] But I think Boolean should be fine [07:51:03] so, it would be something like 6 inputs? Tense, aspect, mood, person (number), person (masc/fem), person (pl/sg) [07:51:04] and you would use strings for TAM? [07:53:00] plus the mutations of course, so 7 inputs... [07:56:22] unless someone is willing to help, I think I'll work on something easier and more useful than conjugation (and I don't want to be the first, I'd like more examples beforehand) [08:57:10] I have been thinking of this question recently, and I am writing up something, but my thought is that we would like probably enumeration types for these features (e.g. grammatical genders, persons, etc.), otherwise things might be too confusing. But I think that's better figured out on-wiki. [09:42:11] that might work for breton, but note that in general grammatical number and grammatical gender are not boolean (re @Feeglgeef: plural is a Boolean and feminine is a Boolean) [09:44:18] exaclty, it won't even work for Breton noun, but it does work for Breton verbs [09:44:19] that's why a specific type/list/whatever would be better (and is probably even required) (re @Nikki: that might work for breton, but note that in general grammatical number and grammatical gender are not boolean) [09:51:23] I would expect separate inputs, and for it to not do anything if you give it an invalid combination. people already have other ways to give functions bad input (e.g. a string that isn't a plausible word, a lexeme id that doesn't exist) (re @Nicolas: which begs the question, is it one or three inputs : the list above or number/plural-singular/masc-fem [09:51:24] the second could lead to...) [09:55:52] joining them would lead to a huge list for some languages (like german verbs have around 30 forms and that's not a lot compared to other languages) [10:04:22] ok [10:04:24] I'd love to see an example (I'm a pragmatical person, I often work by mimicry) (re @Nikki: I would expect separate inputs, and for it to not do anything if you give it an invalid combination. people already have other w...) [10:04:40] yeah, French and Breton both have ~80 forms for verbs (re @Nikki: joining them would lead to a huge list for some languages (like german verbs have around 30 forms and that's not a lot compared ...) [10:05:04] and still, some languages have more ;) [10:11:55] speaking of mimicry and consistency, how should we understand Z6006 and Z6040 exactly? [10:11:57] is is the sens/rank of *a* lexeme/statement or of *the* lexeme/statement? (in English it's not marked - and implicit ? - but in other languages the article is explicitly needed) [11:38:54] Z6040 is the enumeration of possible ranks (i.e. it has three values, preferred, normal, and deprecated) [11:40:08] Z6006 is used to represent the sense on a Lexeme (similar to how Z6004 is a form on a Lexeme) [12:03:18] I would say Z6040 is *the* rank of a particular Wikidata statement and Z6006 is *one* (“a”) sense of the lexeme. [12:21:50] but it's not any sense, it the sens requested, so definite, right? (re @Al: I would say Z6040 is *the* rank of a particular Wikidata statement and Z6006 is *one* (“a”) sense of the lexeme.) [12:44:03] In general, it is any one of the many senses of the (requested) lexeme. In the context of Z6826, you are specifying which one of the senses you require by providing its identifier, the identifier of one particular sense (the required sense, if you like). So I think it depends on the context (unless there is only one sense). (re @Nicolas: but it's not any sense, it the [12:44:03] sens reques [12:44:04] ted, so definite, right?) [14:25:42] I created a phabricator task for this (re @Nicolas: exaclty, it won't even work for Breton noun, but it does work for Breton verbs [14:25:42] that's why a specific type/list/whatever would be...) [14:25:44] Sometime [14:27:39] My idea is that we have dropdown options for functions that don't require a separate type [14:28:34] This would work for things that don't require convenient data transfer [21:44:59] I think that maintainers should have more permissions (delete in mainspace, create objects of any type, change objects of any type) (re @vrandecic: I see two conditions before giving out maintainer rights: 1) having a more robust community, 2) having an updated type editing e...) [21:55:54] I think it's preferable to have few maintainers that have a really good understanding of Wikifunctions, the function model, and ideally all supported programming languages, with many permissions, instead of many that all have a somewhat good understanding and limiting the permissions of the group to account for that