dholbach | davidcalle, mhall119: do you know if we have any deployment blockers right now? | 14:13 |
---|---|---|
davidcalle | dholbach: afaik no, but that's what I'm looking into, I'm reviewing the mojo spec, especially where it pulls data from | 14:14 |
dholbach | cool | 14:14 |
dholbach | thanks a lot for looking into that | 14:15 |
davidcalle | I know there is at least one backport I need to do from the on-server branch to the lp branch (cause you can't actually push branch changes from the server) | 14:17 |
dholbach | ok | 14:31 |
mhall119 | davidcalle: how are we going to manage the staging branch from now on, since the the way we had been using it isn't going to work | 14:33 |
davidcalle | mhall119: I would be in favor of ditching it and going back to feature branches merging in to trunk, then prod. From there testing on staging, then real prod. My very-short tem goal is to make a spec allowing a local deployment. | 14:35 |
mhall119 | davidcalle: in that case, I'd say our branch merging workflow should mimick our deployment workflow, trunk->staging->production | 14:43 |
mhall119 | which would make the staging environment a little less flexible, as we will need to coordinate deployments to it | 14:43 |
mhall119 | but, if you get local deployments working with the spec, that becomes a non-issue | 14:44 |
mhall119 | so staging become "pre-production" instead of "testing" | 14:44 |
davidcalle | mhall119: it appears to be our only option (without touching the current juju/mojo processes), yeah | 14:45 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!