These are chat archives for openworm/ChannelWorm

2nd
Jul 2015
Travis Jacobs
@travs
Jul 02 2015 20:34

@VahidGh @miladjafary
As we've discussed before, the data model at this stage is constantly in flux. For this reason, it's hard to write adapters that connect the data model of ChannelWorm (a moving target) with data models from PyOpenWorm (a more stable target).

The reason for PyOpenWorm's stability right now is because of the way we isolate changes to the codebase in a dev branch. We only push to master when these features/patches are complete, and everyone agrees that it's time to push.

I propose that we adopt a similar scheme for ChannelWorm, and push all of our changes from patching branches to a dev branch first, then merge dev to master when we agree that the changes are significant enough to merit updating the web-app (which reads from master).

This will mean master is updated less frequently, so the web-app will be updated less frequently as well, which means it will also be more stable.

If you guys agree, we can start merging all our changes into the dev branch I created; I've already merged some of my own changes in there!

Vahid Ghayoomie
@VahidGh
Jul 02 2015 20:52
@travs, I agree with the dev branch idea, but how should we agree on the right time for merging, by agreeing first on a meeting? :D
Vahid Ghayoomie
@VahidGh
Jul 02 2015 21:08
As for the instability for working on data models, I suggest postponing this to a time that we are going to start data entry and it is more stable in the CW side
Travis Jacobs
@travs
Jul 02 2015 21:53

@VahidGh
Ok, postponing the adapters you mean? That makes sense.

And yes, I think since we're in a rapid development phase right now, agreeing during a meeting to push to master is more than enough.
Even just creating the dev -> master PR when one thinks the changes in dev are good to merge would be sufficient, and we can discuss whether to merge inside the PR comments.

So as I understand it now, when the authentication is in place we should be able to start adding to the database, and thus start the curation process. As soon as this is possible, @VahidGh, @stripathy and I should meet to start nailing down a workflow for this process (probably this weekend or early next week).
Joe Bowen
@joebowen
Jul 02 2015 23:54
@all, I just wanted to let everyone know that work is getting extremely busy and I won't be able to dedicate the time to this subproject. If there's anything needed on my end with regards to moving anything over, feel free to ping me. :-)