[14:46:55] [1/2] https://cdn.discordapp.com/attachments/1006789349498699827/1519715518854992063/image.jpg?ex=6a3e90de&is=6a3d3f5e&hm=aef2ab5cdb8ef87ea12ec699aa89022e2a680ec2f00455fd48f713e596199f25& [14:46:55] [2/2] https://cdn.discordapp.com/attachments/1006789349498699827/1519715519492788438/image.jpg?ex=6a3e90de&is=6a3d3f5e&hm=3aaa484e02bb3e13fa1c7174caef7f3246aa8c2a3913e81a0d435c02f4c0fca5& [19:00:35] If we remove this from our robots.txt Bing should respect our special page rules. Alternatively we dump every rule we have for `*` into the `bingbot` section as well. [22:24:55] For 1.46 extension testing, is it better to have one paste for all backtraces, or individual pastes for individual backtraces? [22:40:34] [1/2] [22:40:34] [2/2] Does this mean our local patches also have to be applied to REL1_46 then? [23:30:58] IMO ideally one task per backtrace [23:31:29] They should be automatically applied, as long as they're in mediawiki-repos [23:36:13] the patch is in mediawiki-repos and is intended to apply to all branches [23:37:08] It's not applied, I guess it was somehow removed... [23:37:21] should be applied now though [23:39:09] yup fix works now [23:40:45] Not sure why we've started creating pastes btw [23:40:48] Those are harder to track [23:41:17] And with subtasks it's easier to block other tasks [23:42:23] 🤷‍♀️ that's what we did last time. Tasks do sound more sensible so I'll do that from now on. [23:46:57] Last time I think we had one paste for all errors, which at least made it easier to track whether there are still unresolved errors