mozilla :: #balrog

10 Aug 2017
13:26bhearsumg'day folks
13:26bhearsumnice to see you bhavishya_
13:59cloudops-ansiblebalrog-admin #176: building master-5504b2a525d388d15406f8c51c85090678c6155c
14:10cloudops-ansiblebalrog-web #146: mozilla/balrog:master-5504b2a525d388d15406f8c51c85090678c6155c deployed to stage /cc relud bhearsum
14:12cloudops-ansiblebalrog-admin #176: master-5504b2a525d388d15406f8c51c85090678c6155c deployed to stage /cc relud bhearsum
14:17aksareenbhearsum: I was trying to find the issue with UI sending {data: null/object} and not type: string for sc/releases/sc_id API . In both update and delete change_types this method is invoked from the service (https://github.com/mozilla/balrog/blob/master/ui/app/js/services/releases_service.js#L102) which is invoked in the controllers at this line
14:17aksareen(https://github.com/mozilla/balrog/blob/master/ui/app/js/controllers/release_scheduled_change_edit_controller.js#L99) . The $scope.sc contains the object data model (https://github.com/mozilla/balrog/blob/master/ui/app/js/controllers/release_scheduled_change_edit_controller.js#L7) but I cannot seem to find how it is getting set/passed from any other method.
14:25bhearsumhave you tried adding some console.log() or alert()'s in various places to help determine when it gets added?
14:25bhearsumit's possible that it exists as soon as the modal is loaded, since $scope.sc is initialized by copying whatever is passed in
15:23ex-devGood day for all!
15:36ex-devHey bhearsum, I saw that hypothesis pkg was merged today.
15:36ex-devI will keep queryversion=1, edge case ok?
15:41ex-devI will keep queryversion=1, edge case ok (as hardcoded test)?
15:58aksareenbheasrum: cannot seem to log it. I've edited this line(https://github.com/mozilla/balrog/blob/master/ui/app/js/controllers/release_scheduled_change_edit_controller.js#L5) to log to the console but cannot seem to see it anywhere . Here's the console : https://pastebin.mozilla.org/9029399 . What am I doing wrong ?
16:26bhearsumex-dev: i think there's a way to ensure a particular value is always tested
16:27bhearsumbut you can keep it as a separate test too, i don't feel strongly
16:28bhearsumaksareen: where are you looking for it?
16:28aksareenbhearsum: in the browser's console log
16:29bhearsumstrange
16:29bhearsumcan you paste the actual code you're trying?
16:36aksareenbhearsum: oh wait. got the output in the logs. Here's the log for $scope.sc before and after the request. https://pastebin.mozilla.org/9029407
16:37aksareenIt seems the entire release_scheduled_change is fetched from the GET operation and loaded into the UI and then edited field is sent back along with entire model
16:38aksareenSo it is fetching the entire row and updating the entire row
16:38bhearsumthat's not too surprising given the copy that happens at https://github.com/mozilla/balrog/blob/master/ui/app/js/controllers/release_scheduled_change_edit_controller.js#L7
16:43aksareenbhearsum: am confused on what to do next ?
16:45aksareenwhere should I add statement to delete $scope.sc.data field if $scope.sc.change_type == "delete" ?
16:48aksareenmy bad. It should be in this method : https://github.com/mozilla/balrog/blob/master/ui/app/js/controllers/release_scheduled_change_edit_controller.js#L70
16:50bhearsumyeah, that's what i was going to suggest
16:53aksareenbhearsum: yay , it's working now
16:53bhearsumwoohoo!
18:01ex-devbhearsum: I made changes to PR #370.
18:13bhearsum\o/
18:13bhearsumi'll have another look at PRs before i head out today
18:42aksareenbhearsum: i'm seeing import error "ImportError: No module named hypothesis" when executing tests via tox. I've updated my virtual-environment too and installed all packages from requirements-*.txt files. Still error isn't going away
18:44bhearsumyou're running with tox directly?
18:45bhearsumor with run-tests.sh?
18:46aksareentox
18:46aksareentox via a virtualenv
18:47bhearsumok, i don't think tox recreates the virtualenv by default
18:47bhearsumso you'll need to install hypothesis into it, or get tox to recreate it
18:50aksareenalready installed hypothesis in virtualenv. used the command pip install -r requirements.txt/requirements-tox.txt/requirements-test.txt and then ran the tests
18:50aksareenbhearsum: ^
18:53bhearsumdoes it work via run-tests.sh?
18:54bhearsumyou may want to try deleting .tox and trying again too
18:54bhearsumnote that tox itself uses a different virtualenv, that's stored in .tox
18:54bhearsumso just running "pip install" won't do it
19:02aksareenbhearsum: updated the PR. All Test cases passed. Still working on fixing this in my local environment
19:02bhearsumgreat
19:02bhearsumi'll have a look in a bit
19:06ex-devaksareen, I did it today to force recreate local, try "tox -r -e py27 -- auslib/test/web/test_client.py"
19:08aksareenThanks ex-dev it worked
19:08bhearsum\o/
19:11ex-devGreat! -r recreate the env, no need use again
19:15aksareenbhearsum: let's hope this final one gets merged today *fingers crossed*
19:16bhearsumagreed :)
20:29bhavishya_bhearsum: same here;)
21:02aksareenbhearsum: commented on the PR. Updating it in 10 min
21:03bhearsumk
21:03bhearsumi have to head out now, i'll have another look tomorrow
21:04cloudops-ansiblebalrog-web #147: building mozilla/balrog:master-2bcdd4c115e89a61eb1d09b66aea3300ad803f30
21:13cloudops-ansiblebalrog-web #147: mozilla/balrog:master-2bcdd4c115e89a61eb1d09b66aea3300ad803f30 deployed to stage /cc relud bhearsum
21:16cloudops-ansiblebalrog-admin #177: master-2bcdd4c115e89a61eb1d09b66aea3300ad803f30 deployed to stage /cc relud bhearsum
11 Aug 2017
No messages
   
Last message: 67 days and 23 hours ago