On 2016-05-03 08:45:40 +0200, Oswald Buddenhagen wrote: > On Mon, May 02, 2016 at 02:41:54PM +0200, Vincent Lefevre wrote: > > What I've said is that the work is done in a separate branch (a feature > > branch). Then, all that needs to be done is a merge to the main branch > > seen as a single commit (e.g., for git, "git merge --no-ff"). That way, > > one doesn't see the dirty work done in the feature branch, just the > > merge. > > > you're describing a squash merge. that's also not what we want - we want > *series* of clean patches (except in rare cases where a single atomic > commit is indeed sufficient).
I don't think that a *series* of patches is useful, at least for the integration in the main branch. The use of window structures should be seen as a single feature as a whole. Thus a single commit should be fine. -- Vincent Lefèvre <vinc...@vinc17.net> - Web: <https://www.vinc17.net/> 100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/> Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)
