[phobos] [D-Programming-Language/phobos] a15e68: Accidentally committed with a bunch of debugging c...

[Michel Fortin]

Le 2011-05-04 à 11:06, Steve Schveighoffer a écrit :

>> From: Andrei Alexandrescu <andrei at erdani.com>
> 
>> It's difficult to define the "trivial" category. Again, at work we did have the experience that something was considered trivial but caused breakages when deployed. Anyhow, probably we're not at the size we can afford to review each change.
> 
> The Phobos community could be huge, and I still wouldn't have time to review lots of changes.  I've only reviewed one, and that was because it was a pull request changing my code (RedBlackTree).  Not saying that quantity of developers doesn't help, but certainly there are other factors.
> 
> I'd rather have the pull requests reserved for things that are complex or new, and leave small bug fixes to straight commits.  Requiring a review for everything is overkill, and actually counterproductive. Why can't we have simple guidelines that can be overridden when needed?  The only hard rule is, don't break the build.  If it breaks the build, it needs to be reverted.

I'd like to make a suggestion.

Perhaps it'd make sense that many small things could be aggregated in a single pull request and reviewed together. This way the reviewer wouldn't have to merge/build/unittest each commit separately, it'd be done in bunches. All you need is to maintain a branch for those small things, and issue a pull request for that branch every time you feel it contains enough for a review.

-- 
Michel Fortin
michel.fortin at michelf.com
http://michelf.com/