[15:17:04] clarakosi: hey! quick check-in: do you feel you have enough info on the test runner stuff? [15:17:37] do you feel you have a good grip of what we'll be doing? do you know where to start? [15:17:53] I want to make sure you don't fell left hanging in the air [15:19:37] duesen_: yeah I get the gist of it. I spent some time looking at service-checker yesterday. And I'm assuming I'd be starting with T219883 [15:19:38] T219883: Specify file format for API test definitions - https://phabricator.wikimedia.org/T219883 [15:21:41] clarakosi: yes, if you feel up to it. I'd suggest to start with what x-ample does, and add to it as described in the ticket. [15:22:03] If you get stuck or need someone to discuss ideas with, ping me. [15:22:16] if I don't respon on IRC, shoot me an email. [15:23:23] duesen_: alright! And I should of probably asked you this at the meeting but when are you online? [15:23:59] working hours that is [15:32:59] clarakosi: varies wildly by day to day. My core hours are more or less 10am to 8pm CET/CEST. But I'm often online after tha, and often off for a few hours in between. [15:34:09] The most reliable time to catch me is probably past 6pm CET/CEST [15:35:09] On Mondays I'm the most flexible [15:35:56] Alright [16:01:35] Krinkle: Almost everything you said at T219974#5082048 seems like it would also apply to not showing deletion logs at the top of pages like https://en.wikipedia.org/w/index.php?title=Category:Articles_with_unsourced_statements_from_January_2007&action=edit&redlink=1 (although I doubt anyone would gloat over that specific example). [16:02:33] anomie: True, but we only show that to users able to perform the edit action there. [16:02:36] so logged-out doesn't see it. [16:03:29] But yes, it's a complicated area. There'll be a different trade off depending on the specific feature/case/impact/usage patterns. [16:03:42] Krinkle: Which seems no different from trying to shorten a URL that was already deleted, since if you don't have the ability to create shortened URLs you presumably can't get the "already deleted" error either. [16:04:06] anomie: Yes, but metadata vs content. [16:04:20] anomie: We don't expose the content of articles through a "this content already existed previously" [16:04:45] we could consider upon resolving a deleted short url state "It was deleted." instead of "Not found" [16:04:46] Krinkle: The target of the redirect isn't really "content" either. [16:06:43] I'm not going to discuss the details on IRC. It's a security task. There is something worth protecting. Whether we classify it as content isn't the point. [16:10:20] I'd argue against "There is something worth protecting." But, meh.