[06:53:12] hi Petra, are there any updates on this [07:12:43] It should have been fixed last night [07:13:25] Oh you're running your own [07:13:43] Please set a compliant UA and then let us know and we'll exempt it [07:17:48] Is the bot exemption done purely on UA? In all honesty I'm expecting MH to be more aggressive on requests coming from Azure/GCP/AWS services, which unfortunately affect GitHub Actions/GitLab Runner CI/CD pipelines. In that event it would be good to know how strict/draconian the expected measures will be. [07:18:33] forwarding this question as well which remained unanswered (and yes I'm aware spoofing either is trivial by bot) [07:18:49] It will depend on exactly what block you're hitting as to what options we have available [07:19:17] I don't think we've used Auth headers before though [07:20:01] are there different blocks? I just assumed it'd be the same 403 message outputted on the job log [07:20:39] There are lots of different blocks yes [07:20:54] Some on IPs, some on UAs, some on ASNs, some on patterns of behaviour [07:22:13] I meant like an OAuth token passed onto the Auth bearer header. Having my request from GitLab runner (GCP) blocked when I was correctly providing OAuth token and UA was strange at the time. [07:24:05] oh you meant information that is only known on the tech side and wouldn't be passed onto the client side 😦 [07:28:00] I know you mean, I just don't think we've used it as a factor in deciding whether or not to block someone yet [07:28:46] We are unlikely to tell people exactly what we use for every single block or exemption otherwise it'd be a bit pointless [07:29:04] If you are blocked and need help getting unblocked, email tech@ or DM me and we will investigate [07:29:24] If you DM me, expect to be waiting a few hours at least though as I'm about to go out [07:29:46] no I'm good, I'm just pre-empting for a possible scenario [10:18:52] If I want the DOORS Wiki to be a repository for multiple interlanguage wikis, can I just manually set a FileRepo for each of these wikis in the `mw-config` repo? [10:21:49] You can do it from ManageWiki [10:22:12] There's some shared file repo stuff in there [10:24:14] But doesn't that only let me set one wiki per value? [10:25:17] It should work, it's been done before. Try what you can do and file a task for the rest. [10:25:26] Alrighty. [10:25:48] One wiki to share files of, one acts as a file repo to everything else [10:25:55] You can't have wikis share files both ways [10:26:29] Yeah I know that, I just want one wiki to be a pseudo-commons wiki, while it also has its own content. [10:26:43] That way we don't need to reupload files everytime we make a new interlanguage wiki. [10:29:56] That's fine [12:51:51] How long does an account have to wait to make a new wiki request after already making one:? [12:52:05] <_arawynn> 3 days [12:52:22] Ughghghghg [12:52:39] I'm trying to set up interlanguage wikis for the DOORS Wiki, and I don't wanna wait three days each time to make a new request for them. [12:57:06] could plot it out and do initial setup on each in sequence [13:02:28] Special:RequestImport is failing on a 76.9 MB file. [13:10:14] as far as I'm aware this wouldn't apply to anyone with `noratelimit`, i.e. meta admins/bureaucrats and various global groups [13:11:08] [1/4] https://cdn.discordapp.com/attachments/1006789349498699827/1546145727858741338/image_f45h39.png?ex=6a9eb7eb&is=6a9d666b&hm=8833739fe50a5097d97b4c1ab9d63454f1c8d3e8b7688fba1d6ba04e801491d6& [13:11:08] [2/4] https://cdn.discordapp.com/attachments/1006789349498699827/1546145728387358832/PXL_20260822_133855_289.png?ex=6a9eb7eb&is=6a9d666b&hm=f32ed1aca2743b819d505928127ef5d9ce6a26fe4eb31e4cccd33d6c65b903b6& [13:11:08] [3/4] https://cdn.discordapp.com/attachments/1006789349498699827/1546145728890671134/20260815_013921.png?ex=6a9eb7eb&is=6a9d666b&hm=8885f272dcb7c42e7cfedb6ac069aa3632e0c5f3fc29b157496e6971951f0622& [13:11:09] [4/4] https://cdn.discordapp.com/attachments/1006789349498699827/1546145729288998962/IMG_8195.png?ex=6a9eb7eb&is=6a9d666b&hm=3477c8a484f40fcd2e0221ed0de16962991940755c2d6125ea8757b461b52210& [13:15:48] So, set `$wmgSharedUploadClientDBname` to `zhdoorswiki ptbrdoorswiki`? [13:24:50] Yeah, I don't think I can set two wikis in that value [13:31:43] And I can't for the life of me get this requestimport working. [13:38:36] Open a task to get that to support multiple wikis then pls [13:39:04] Alrighty, I will. [13:59:15] Yeah, but WR isn't a global role, so I'm not excused via the AF. [16:37:48] How many languages are you looking at doing? Stewards can help get that going on your behalf if you have a bunch ready to go. [16:41:02] Anything over 25mb-50mb is probably going to be trouble and need to go the phorge route [16:43:43] Yeah, I made a phorge ticket. [16:47:07] Uhhh, we're getting requests increasingly, so far I just need a French Wiki created, but we also might be getting a Hindi wiki soon. [16:54:57] Got it. If you end up in a situation where you're rate blocked, let us know and we can assist in getting the bonus ones created. [16:55:33] I'll also chat about possibly adding a rate limit exemption to WRs, but that may have some unforeseen consequences/rights issues I need to discuss with the rest of the team. [21:37:35] given a bit of policy railing on self creation I wouldn't mind too much [21:38:21] I feel but do not enforce or seek to actively establish that reviewing one's own request is not optimal practice [22:01:30] [1/2] The wiki reviewers page says that focusing primarily on approving one's own requests over others' is grounds for steward revocation, but as far as I know that's as close as written policy gets. [22:01:31] [2/2] But ye, it seems improper to review your own requests. That's just a conflict of interests. [22:11:40] It's a rule that exists because someone made it necessary, as most of them are. [22:13:14] yes there was a very specific example there [22:13:33] the process has since made it very improbable [22:15:24] Welllllll, there are still ways [22:15:32] But yes, improbable [22:25:55] well you know I see all kinds of tricks but would still like to believe they're improbable [23:25:59] I have tested it to working up to 300mb so that is very odd. [23:35:07] that was a 7z file that aeywoo tried to upload [23:35:17] I find it gets increasingly iffy on anything above 50-100 and prone to terminating early [23:35:52] Oh, yeah, if it's expecting xml and gets a .7z, that'd do it too [23:36:20] It won't let you upload anything but XML. [23:37:14] My understanding might also be dated [23:37:19] oh yeah, the 7z file is only 1.7MB [23:37:31] I am going to be working on expanding it to image imports + chunking and maybe extracting archives though soon also. [23:38:19] With chunking it would allow it to work with massive imports also.