Hi, Le dim. 7 mai 2023 à 17:54, Konstantin Kharlamov <hi-an...@yandex.ru> a écrit :
> On Sun, 2023-05-07 at 17:21 +0200, Nate Graham wrote: > > There's also no reason anymore why they need to use a work branch in the > > main repo; a fork works just fine. I do nearly all of my development > > using personal forks; it's a 100% supported first-class citizen > experience. > > +1 > > Ok for me to go with forks for GSoC but as Ben said, we probably need to find something about abandoned forks. I remember when we were integrating Gitlab at work, people for some reason > were > attempting to create branches in the upstream repo rather than on their own > fork. Had to do some explanation that this isn't the way to go, because > after an > year of such workflow upstream will end up with hundreds if not thousands > of > abandoned branches that serve no purpose, and who's gonna clean that up > 🤷♂️ > Especially given that people may come and go, and so in some cases you > wouldn't > even find an author of a branch (and in some cases it's hard to even > figure out > who could be the author in the first place). > The issues you mentioned are the same with forks. Instead of having abandoned branches in the main repo where maintainers could know if they are active or not, we will have abandoned branches in abandoned repositories without MR to know they exist and they would need to be cleaned. Cheers, Johnny