Re: On pull request workflows for the GNU toolchain

[Jonathan Wakely via Gdb <gdb@sourceware.org>]

On Mon, 23 Sept 2024 at 19:00, Eric Gallager via Gcc <gcc@gcc.gnu.org> wrote:
>
> On Mon, Sep 23, 2024 at 8:09 AM Thomas Koenig via Gcc <gcc@gcc.gnu.org> wrote:
> >
> > [For the fortran people: Discussion on gcc@]
> >
> > Just a general remark.
> >
> > There are people, such as myself, who regularly mess up
> > their git repositories because they have no mental model
> > of what git is doing (case in point: The Fortran unsigned
> > branch, which I managed to put into an unrepairable state
> > despite considerable help from people who tried to help me
> > fix it).
>
> As one such person who has messed up his fork of GCC, I'd just like to
> note that in my particular case at least, I messed it up because I was
> trying to apply GitHub's model for git usage, while the GCC project
> has a very different model for git usage, and the two don't exactly
> play very well with one another.

Only in the sense that the GCC project wants a linear history without
merge requests, and GitHub will happily let you create whatever mess
of merges and jumbled commits you choose to. Using pull requests won't
change that - you will need to sort your branch out before it will get
merged.


> I see switching to a pull request
> model as reducing the chances of people getting their forks into
> unusable states, rather than increasing it.

I don't think it will change it at all, you'll still be able to do
anything to your fork, it just won't get approved for merging if it's
a mess.