[16:59:39] <Pchelolo>	 mdholloway: how's the vary and accept-language patches going? Any change we can deploy those soonish?
[17:01:53] <mdholloway>	 Pchelolo: lol, i've just -1ed https://gerrit.wikimedia.org/r/#/c/mediawiki/services/mobileapps/+/441320/, but if James_F makes a quick update then i think we can get it deployed today
[17:02:05] <mdholloway>	 i was planning to do a deployment for sure
[17:02:29] <Pchelolo>	 awesome. we ave RESTBase counterpart ready-ish, just some lines in the tests commented out
[18:27:20] <mdholloway>	 Pchelolo: does yours depend on ours being deployed?  there's some tests i want to update but they depend on prod restbase emitting 'vary' and 'content-language' headers
[18:27:24] <mdholloway>	 lol, maybe we're in a standoff
[18:27:42] * James_F grins.
[18:28:00] <Pchelolo>	 mdholloway: no, it's doesn't, we don't enable it for mobile and summary yet, only our tests depend on it
[18:28:58] <Pchelolo>	 we're waiting for Parsoid to be deployed to deploy ours
[18:30:49] <James_F>	 They're doing that right now, I think.
[18:30:50] <mdholloway>	 James_F: the thing that's troubling me is that the update to test/utils/headers.js could mask an invalid absence of a vary or content-language header
[18:31:11] <mdholloway>	 also, undefined !== 'undefined'
[18:31:36] <mdholloway>	 for the first problem, maybe we just need a TODO
[18:31:39] <James_F>	 mdholloway: Yeah, I was particularly testing for the header getting added as a string cast of undefined because I caused that error in my early work.
[18:32:16] <mdholloway>	 ah, maybe that check is worth leaving, then
[18:32:17] <James_F>	 mdholloway: In general, the tests for headers are totally insufficient for the codebase before my patch, as they are too generic to be specific to particular responses.
[18:32:40] <James_F>	 Other than "expected header Foo to be present and it wasn't".
[18:32:49] <mdholloway>	 ha, fair enough.
[18:33:12] <James_F>	 We don't even check that headers we didn't expect to be present were. :-) Definitely some TODOs around this.
[18:33:51] <mdholloway>	 yeah, let's phab it.
[18:34:12] <mdholloway>	 MCS was much smaller when that header testing approach came about.
[18:41:28] <James_F>	 mdholloway: Made T198134
[18:41:28] <stashbot>	 T198134: Improve per-endpoint response header testing in MCS - https://phabricator.wikimedia.org/T198134
[18:41:44] <mdholloway>	 James_F: great, thank you!