- 2020.03.24 13:33 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Frank Warmerdam
- 2020.03.24 14:21 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Bob Friesenhahn
- 2020.03.24 15:07 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Leonard Rosenthol
- 2020.03.24 18:45 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Roger Leigh
- 2020.03.25 15:42 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Olivier Paquet
- 2020.03.26 09:21 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Even Rouault
2020.03.25 16:18 "Re: [Tiff] Motion to grant push rights to Thomas Bernard", by Olivier Paquet
Which makes me wonder: do we have some policy on merge requests? eg. should we wait for review/comments from someone else before merging some code? Wait some amount of time? Just merge and hope for the best? I don't think we had that discussion when we switched to gitlab. As far as I know, CVS worked more or less as "discuss anything uncertain on the mailing list and please don't break stuff when you commit".
My personal opinion regarding this would be to have a soft policy:
- use of merge requests should be the norm, and waiting for CI tests to pass (@Roger wondering if it would be possible to have the Windows tests to be run on merge requests as well ?)
- depending on the complexity/risk of the change and confidence you have in it, wait a bit for potential feedback. And for things that might require feedback from a wider audience, ask on the mailing list. Not sure if we can express that in less subjective terms...
That seems reasonable and pretty much what is being done. I don't think we need hard rules at this point, just some guidance.
I know OSS projects which require approval by other members for any merge request to be merged, but there are significantly more staffed than libtiff.
Indeed, given that there are not that many of us and even fewer with spare time, it would be difficult and likely result in a large backlog.
Olivier