UbuntuIRC / 2020 /05 /22 /#launchpad-dev.txt
niansa
Initial commit
4aa5fce
=== wgrant_ is now known as wgrant
[13:06] <roadmr> every time the "propose an mp here" remote message saves me time, I'll use it to come here and sing the Launchpad team's praises 😄 (or until you tell me to shut up and go away :)
[13:07] <roadmr> oh also a question: has it ever been possible to log in using username instead of email on Launchpad or SSO? if so, any idea why that was changed to using email only?
[13:15] <cjwatson> I saw a few upset SSO questions about that
[13:16] <cjwatson> If it ever was in LP I think it was before my time
[13:16] <cjwatson> But honestly I'm not sure it ever occurred to me to look
[13:19] <roadmr> cjwatson: yes, people caring about that minute detail tend to be the upsetty kind :/
[13:19] <roadmr> "please don't assume I know my email address" say WHAT
[13:20] <cjwatson> I can sort of understand that from the "generate unique email address for each different thing" crowd
[13:20] <roadmr> well yes but if you're in that crowd I'd expect you to have proper tracking of all the addresses you're spitting out I guess
[13:21] <roadmr> cjwatson: I was thinking it might have happened in the SSO split - when Launchpad was in charge of logins, a username was mandatory and could be used to log in. SSO didn't require a username for a long time, so the email was the only consistent login option.
[13:21] <roadmr> These days all new accounts do have a username
[13:21] <cjwatson> It is possible, but all this was before my time and I've never done the archaeology, I'm afraid
[13:22] <cjwatson> I can't immediately think of a reason why you shouldn't be able to log in by username. It seems an odd omission.
[13:22] <cjwatson> (Assuming you have one)
[13:22] <roadmr> cjwatson: maybe we can ask the chief archaeologist in a couple of hours
[13:23] <cjwatson> The theory that it was something to do with SSO not having usernames for a long time seems a likely one.
[15:41] <tomwardill> cjwatson: if I wanted to add the nodejs from u1hackers to LP, is there anything special I'd have to do to test it? (I've run a LP make cycle locally and it seems happy)
[15:44] <cjwatson> tomwardill: Check that CSS and JS works in a browser after make; run tests for --layer=YUITestLayer --layer=YUIAppServerLayer
[15:45] <cjwatson> tomwardill: And we'd need to not land anything that depends on it until relevant machines (I guess carob, banana, nutmeg, atemoya, labbu) have been upgraded (which you can check in Landscape)
[15:46] <cjwatson> Indeed upgrading would require a ticket to copy it into the relevant IS-owned staging and production PPAs
[15:46] <cjwatson> https://launchpad.net/~canonical-is-sa/+archive/ubuntu/launchpad-staging and https://launchpad.net/~canonical-is-sa/+archive/ubuntu/launchpad
[15:46] <tomwardill> right
[16:28] <tomwardill> verbose 3.205 Performing "GET" request to "https://registry.yarnpkg.com/node-sass/-/node-sass-4.14.1.tgz".
[16:28] <tomwardill> error request@2.88.2: The engine "node" is incompatible with this module. Expected version ">= 6".
[16:29] <tomwardill> boo, okay, so we definitely need the SCA node backport
[16:32] * cjwatson nods
[16:50] <tomwardill> okay, I'll start the upgrade dance on Monday I guess
[16:50] <tomwardill> no wait
[16:50] <tomwardill> Tuesday
[17:19] <cjwatson> pappacena: thanks for that MP, dinnertime here now but I'll try to have a quick look at it this evening
[17:19] * pappacena was about to send the link here :)
[17:20] <pappacena> Thanks, cjwatson. Whenever you have time. If you find something while reviewing, I'll probably still be around to make adjustments.
[17:21] * pappacena going back to OCI project / project stuff for now
[17:28] <SpecialK|Canon> I just realised the views are live!
[17:28] <SpecialK|Canon> Nice work
[17:28] <SpecialK|Canon> Let's get that in next week's weekly
[17:31] <pappacena> :-) and soon we will have those views for project based oci project project project too! hehe
[17:38] <tomwardill> Sure you missed a ‘project’ in there.
[17:38] <tomwardill> we can find room for more.
[17:41] * pappacena 😂
[19:09] <cjwatson> pappacena: looks fine, just one trivial nit
[19:10] <cjwatson> pappacena: (though I was curious why the store.invalidate() stuff in the second test, rather than just testing the object you get back from .inject()?)
[19:11] <pappacena> let me check
[19:22] <pappacena> Ah, the idea was to flush, but I guess it doesn't make any difference since it's fetching from the database after saving. I'll remove it.
[19:23] <pappacena> Fixed the typo. I'll land that MP
[19:51] <cjwatson> Cool, thanks
[19:51] <cjwatson> Should be able to deploy and rerun on Tuesday or so I guess
[19:53] <pappacena> Cool. Hope everything will work then. Enjoy your weekend and thanks for the review!
[19:54] <cjwatson> You too!