[15:59:47] it appears the site in general (if not due to it being locked down) is confusing a lot of people: https://news.ycombinator.com/item?id=36927695 [16:03:53] the scope of our work will be clearer with time, hopefully [18:26:13] I stumbled upon this interesting report which underscores, I believe, the importance of Abstract Wikipedia's vision. [18:26:14] https://amp.dw.com/en/bridging-the-ai-language-gap-in-africa-and-beyond/a-66331763 [18:30:51] I really hope that people won't speak to each other in ZIDs. Like, no one, not even very experienced Wikifunctions users. Is it vain to hope for this? (re @mahir256: it appears the site in general (if not due to it being locked down) is confusing a lot of people: https://news.ycombinator.com/i...) [18:32:03] Q60622784 (re @amire80: I really hope that people won't speak to each other in ZIDs. Like, no one, not even very experienced Wikifunctions users. Is it ...) [18:32:31] Unless, I don't know, there's an exceptionally good reason, like debugging something, and having absolutely no other way to discuss it. [18:32:53] Q93728478! (re @amire80: I really hope that people won't speak to each other in ZIDs. Like, no one, not even very experienced Wikifunctions users. Is it ...) [18:33:38] Q3280607 (re @Jan_ainali: Q60622784) [18:33:46] That's reason #1 why I find discussions about Wikidata tiring. Too may Q and P and L IDs. And linking bots don't make it easier. [18:34:28] You know, when someone says that they don't like something, immediately doing it to them is not nice, not even as a joke. [18:36:05] reducing ambiguity across languages is important tho, and references to identifiers are a good first step [18:36:44] this avoids the bad faith of those who interpret e.g. "the item for Paris" as referring to the Texas town and not the French city for which it was named [18:37:53] Yeah, but a lot of people are very bad at remembering numbers. Most people, I suspect. Even most Wikimedians. [18:38:43] Numbers are a good way to store it internally, but when it comes into conversations, it feels like gatekeeping, even if it's not intentional. [18:39:02] that's why @jhsoby's wikilinksbot is a godsend here; it would be terrible to talk in identifiers where such a facility did not exist (re @amire80: Yeah, but a lot of people are very bad at remembering numbers. Most people, I suspect. Even most Wikimedians.) [18:39:23] and why, I imagine, neither @Jan_ainali or I thought of our responses as 'not nice' (re @amire80: You know, when someone says that they don't like something, immediately doing it to them is not nice, not even as a joke.) [18:40:28] Well, as I said... It actually isn't. It's cool, and it makes it slightly more readable. But only slightly. It's better not to have to rely on it. (re @mahir256: that's why @jhsoby's wikilinksbot is a godsend here; it would be terrible to talk in identifiers where such a facility did not e...) [18:41:16] is this last sentence also applicable to Wikidata itself? (re @amire80: Well, as I said... It actually isn't. It's cool, and it makes it slightly more readable. But only slightly. It's better not to h...) [18:42:53] When people link using templates that show both the label (in any language) and Q/P/L id right in the text, it's far, far more readable. [19:32:28] a large part of the problem is that telegram isn't designed for talking about wikidata entities. on-wiki templates can automatically show you the label you'll see elsewhere in the interface just by giving it the id, but there's no way to do that in telegram. even if you take the time to manually write the label and id with a link, it still won't be the right label for each person reading it [19:40:07] for properties I definitely prefer when people include the ids, because it can be hard to work out which property someone is talking about otherwise, and a label being more readable than an id is not much use if I then can't answer someone's question because I don't actually know what they're referring to [19:42:45] I'm not a fan of the bot though (I don't like that it takes up a lot of space and hides who sent the message) and I try to avoid triggering it (which usually means I'll just paste a link) [19:45:49] The same can be said about the rest of the message. [19:45:50] [19:45:52] Including the ID is always useful, of course. But doing a little something to make sure that more people can read you is useful, too. (re @Nikki: a large part of the problem is that telegram isn't designed for talking about wikidata entities. on-wiki templates can automatic...) [19:46:43] Can bots edit people's messages? I played with the Telegram bot API, but the last time I touched it was probably in 2018. [19:46:54] not as far as I know [19:47:22] I don't think so (re @amire80: Can bots edit people's messages? I played with the Telegram bot API, but the last time I touched it was probably in 2018.) [19:48:32] I'm not sure I'd like wikilinksbot (or really any other Telegram bot) as much if it could modify my messages [20:16:18] of course, but in my experience, the rest of the message has more error tolerance and a small difference in how someone translates something there is not as likely to cause misunderstandings (re @amire80: The same can be said about the rest of the message. [20:16:19] [20:16:20] Including the ID is always useful, of course. But doing a little something ...)