On Sat, Apr 18, 2015 at 5:15 AM, hasufell <hasuf...@gentoo.org> wrote: > On 04/17/2015 07:15 PM, Rich Freeman wrote: >> On Fri, Apr 17, 2015 at 10:41 AM, Alexander Berntsen >> <berna...@gentoo.org> wrote: >>> >>> On 17/04/15 16:33, Andrew Savchenko wrote: >>>> The problem is double effort: previously one developer effort was >>>> needed, now effort is doubled at least >>> You have correctly identified the problem; in order to do things >>> properly one must do things properly, which is more difficult than not >>> doing things properly. >>> >> >> "Properly" is just a matter of requirements. Gentoo has 18k packages >> right now. In my general experience, they install fine maybe 95% of >> the time. >> > > Can you back up your "general experience" with a tinderbox log?
No. Of course, having a review workflow is orthogonal to having a tinderbox. > In addition, you are decreasing "QA" to "compiles". That's not the definition. If it makes you happy s/install/works. It is fairly rare to run into problems with Gentoo packages in my experience. > >> Right now we >> end up dropping packages because we can't find one person to maintain >> them. With a review workflow we'll drop packages if we can't find two >> people to maintain them. > > Nah, that's really not true. With a review workflow there is less need > for actual maintainers! That's the whole point. There is more need for actual maintainers. There is just less need for them to have commit access to the tree. If we instituted a policy that all commits needed to be reviewed it isn't like there would magically be a ton of pull requests headed our way. Users submit patches today, and users would submit patches tomorrow. We're not drowning in them today, and that is unlikely to change. There would still be nobody committing changes to java packages, just like today, and so on. > > I am really confused. I guess some people have never really been in a > different workflow than gentoo to know that it's really not > state-of-the-art. And it really isn't. Not even for distros. > I am not saying that a review workflow is bad. I just don't see how it fixes our actual problems, which is a lack of commits in the first place. You keep using the linux kernel as an example. The kernel has 8 patches per hour and the software is high-complexity. They need a review workflow to vet those changes and filter out the bad ones or get them reworked. Most committers are very motivated to get their code into the kernel. Other distros have MUCH larger userbases and active maintainer communities. They are also much simpler since they don't support mixing and matching random combinations of gcc, libfoo, and so on. Gentoo just doesn't have the same volume of incoming work. Now, if you're talking about making it easier to submit patches, having automated testing, and so on, I'm all for that. There are some already working on that and it will likely become more integrated into the core workflow when we migrate to git. Users can already submit pull requests using the github mirror. We just don't force everybody to do it that way. -- Rich
