devkulkarni: i was looking through the error and the codes for other dashboards, it just seems like the jenkins environment itself checks for the earlier version of pbr in the case of solum
other dashboards have also just implemented just pbr>=1.6, like i did, and it even works in my environment if i change my tox requirements like i mentioned yesterday
devkulkarni
pt_15: I see.. in that case I would recommend pinging folks on the openstack-infra irc channel to understand why the gate failure is happening. you can explain to them that above point
yesterday I was looking at the cli to figure out if we can remove plans from the cli code
essentially the OldAppCommands from solum.py in the cli
vijendar
ok
devkulkarni
I realized that there is the whole githubAuth stuff in the OldAppCommands
I saw that we have some of that in the AppCommands
I remember you had last worked on it
do you know if we have completely moved over the current functionality from OldAppCommands to AppCommands as far as github integration is concerned?
vijendar: also, btw, yesterday I was telling pt_15 that you had attended a workshop on building angularjs plugin with horizon.. I was wondering if there were any tutorials/links etc. that the workshop organizer shared with the attendees?
vijendar
devkulkarni: they promised to share the presentation and example code. I will check if they have sent any email
devkulkarni
vijendar: btw, currently when we deploy an app, the state of the workflow is 'READY' .. I remember you were thinking of changing this to 'DEPLOYMENT_COMPLETE'
do you have a bug for that?
vijendar
devkulkarni: I think did not move the github related code to new app commands
devkulkarni: yes. I think I have that bug with me
devkulkarni
vijendar: oh okay.. that means we cannot yet delete the OldAppCommands
vijendar
devkulkarni: yeah.
devkulkarni
I think we should create a bug to move the github stuff from OldAppCommand to AppCommand
vijendar
sure
devkulkarni: let me know if you want me to create the bug
vijendar: so a general problem with using a special string in the value is that we won't be able to create a languagepack by that name.. this affects any value that we select ('false', 'no-languagepack', 'languagepack-not-provided', etc.). probably we should add another field to the app data model to indicate that a lp is not provided instead of re-using the 'languagepack' field.. what do you think?
vijendar
devkulkarni: you are right. we cannot create languagepack with the special string that we use
devkulkarni
yeah.. let me change that patch to use a new field
I will put -1 on both the patches
vijendar
devkulkarni: but these constants does not seems to be a reasonable name for languagepack
devkulkarni
true.. but it will be very easy for someone to exploit
they may create lp with that name and then their app won't build if they are using that lp
vijendar
what happens if they take advantage of this?
they are shooting in their foot….right?
devkulkarni
basically the app build and deploy won't work for them..
right
vijendar
probably we can have some validation on not allowed languagepack names
devkulkarni
that is another option
if we go the validation route, then I think 'false' is better than 'no-languagepack' or 'languagepack-not-provided'
no?
ppl will understand if we say that you cannot name your lp as 'true' or 'false'
as those strings have special meaning
adrian_otto joined the channel
vijendar
since this is not user visible stuff, I was thinking having no-languagepack as constant makes the code more readable for developers