On 27.04.2014 14:42, Pawel Dziepak wrote:
2014-04-26 5:23 GMT+02:00 Przemysław Pintal <premislaus1988@xxxxxxxxx>:2. Once on the #haiku-pl, pdziepak expressed the opinion, that he likes projects in which the other person, reviews and accepts commit.[...] Last, but not least, there is actually a reason why I didn't proposed such changes to our workflow earlier. While this may work very well with project when the vast majority of developers is employed full-time I am not sure it will work for us with our limited resources.
We tried a strict review policy during the alpha 2 release process. Only patches would be merged to the release branch that had been reviewed by someone other than the author. It didn't work that well. At that time I was working full time on Haiku, including managing the branch. I reviewed most of the patches that I didn't author. It still wasn't easy to find reviewers for all the remaining patches, particularly mine [1].
I think the main reason why a strict review policy cannot work for us ATM is that our number of active developers to size of code base ratio isn't big enough. I suppose one would need at least half a dozen part-time Haiku developers who, together, cover all areas of expertise at least twice (*).
CU, Ingo(*) Or possibly a hundred developers with the average time commitment of the five currently most active developers (not counting Adrien's contract time).
[1] https://dev.haiku-os.org/wiki/R1/Alpha2/MergeTracking
