mozilla :: #conduit

8 Sep 2017
14:43zalunglob: I've checked the way you load patch from S3 in transplant. Lando API is encoding the patch to 'utf-8', saving to the tempfile and uploading fileobj to S3. I'd like to know you would like to have it saved it another way or it's fine as it is.
14:48dklirc or vidyo standups today
14:49davidwalshdkl: I jumped into mars' room and no one was there, so bailed
14:49dklok so irc then i think
14:49zalunI've been there, just without video :)
14:49davidwalshdkl: What did you work on yesterday and where should we go next, in light of what we talked about yesterday?
14:52dklstandup: yesterday i worked on displaying comments in bug reports when a user accepts/rejects a revision. today i was going to play around with your new WIP patch for requiring bug ids for certain repos.
14:56zalunI've been fixing my repo structure, browsed autoland/transplant code to see if Lando API is providing right output, fixed minor difference there
15:28davidwalshdkl: My WIP isn't really applicable anymore, outside of the of the new repo setting
15:28dklah ok so will skip. i was working on a new version of my doorkeeper new revision patch that check for bug id
15:28davidwalshdkl: We'll need to add a "is_bug_required?" check in the validator and the doorkeepr revision, and simply bail if the bug isn't required; that part seems easy
15:29davidwalshIt's the harbormaster part that would be difficult
15:29davidwalshBut with your new patch, I think that problem goes away
16:11davidwalshdkl: Does this need to be an @ instead of a $?
16:13dkldavidwalsh: no since returns a reference to a list, not a actual list
16:16davidwalshOK, cool
16:44davidwalshdkl: Does the subscriber stuff that you're adding in the BMO PR also need to be added in the `sub revision`, or is that the subscriber stuff that's already being done?
17:08dkldavidwalsh: no 'sub revision' does a snapshot of the cc members at that time and adds them as subscribers to the new revision. Thecode in the PR is if the cc list changes
17:08davidwalshBaller, thanks
17:08davidwalshI'm super sensitive about anything that touches the push connector
17:26dkldavidwalsh: if i do 'echo var_dump($data);' in the doorkeeper code, where does it go?
17:26dkldo i need to do $this->mozlog(var_dump($data)); instead?
17:27davidwalshdkl: echo'ing should go to the stdout of the daeman
17:28davidwalshYou don't need to echo with var_dump, btw
17:28davidwalshdkl: I much prefer print_r; that's what I used
18:11davidwalshckolos: We've had the the mozlogging stuff in prod a few days now; can you confirm that those log messages are making it to where you want them to go?
19:19ckolosdavidwalsh: how did they get to prod without a deployment/
19:19davidwalshckolos: Oh, I thought we were doing weekly
19:19davidwalshMy bad if it hasn't
19:19ckolosnot on phabricator, no
19:19ckoloslando, yes
19:19davidwalshCode's been in for over a week, I think
19:19davidwalshckolos: Ahh, OK, my bad
19:20ckolosyeah, no worries
20:12dkldavidwalsh: yes. finally got the magic incantation
20:13dkldavidwalsh: so question. now that i know what to look for, when a bug id is cleared from a revision and the revision was previously private, we set to Public?
20:27davidwalshdkl: I think we need to check with BMO, right?
20:27davidwalshI'm not sure we can assume public
20:28dklyeah but if they clear the bug id. and we poke BMO, BMO will not see a bug id and throw an error
20:30davidwalshOh, hmmm
21:14dkldavidwalsh: well new patch for review with out current behavior so we can modify it later
9 Sep 2017
No messages
Last message: 12 days and 8 hours ago