[postgis-devel] Stable Branch Naming : Git

rmrodriguez at carto.com rmrodriguez at carto.com
Mon Oct 28 10:22:12 PDT 2019


On Mon, Oct 28, 2019 at 6:03 PM Paul Ramsey <pramsey at cleverelephant.ca> wrote:

> As in "push your work to a personal fork and PR from there"? works for
> me (though inevitably I'm going to fat-finger a push to origin)

I'd recommend you to set up the main repo with a different name than
origin. In my case I have `blessed` for the main repo, `origin` for my
fork (where I rewrite history all the time), and then some contributor
forks that I'm interested in. Then you can either create a PR from
your branch, or apply your changes to master and push that into
blessed/master.

On Mon, Oct 28, 2019 at 6:09 PM Greg Troxel <gdt at lexort.com> wrote:
> How are people supposed to read and comment on feature branches (I
> assume that's what you mean) of things that are proposd for merge to
> master?

Same as always, you can create a PR from a fork to any of the mirrors,
or submit a patch manually. Exactly the same procedure that external
contributors follow, which I think is a nice thing.

> I would say it's fine for live feature branches to be in the main repo,
> but branches that are abandoned should be deleted (which is harder than
> it sounds, from experience).

I usually keep 4-5 alive feature branches at all times, and what's
worse I might make short live branches while I test stuff that I then
discard. I don't do a good job of cleaning them, so my Github fork now
contains 368 branches. If all contributors do the same (with 4-5
branches) that's ~30 branches that we need to keep cleaning and that
might get pulled unnecessarily. When I pulled from `blessed` I want
the stable branches and master, not development branches (for that I
can add other remotes).

On Mon, Oct 28, 2019 at 6:09 PM Greg Troxel <gdt at lexort.com> wrote:
>
> Paul Ramsey <pramsey at cleverelephant.ca> writes:
>
> > On Mon, Oct 28, 2019 at 10:02 AM <rmrodriguez at carto.com> wrote:
> >
> >> Regarding this, can we keep only stable branches + master in the fancy
> >> new repo? I don't like having development branches in the main
> >> repository.
> >
> > As in "push your work to a personal fork and PR from there"? works for
> > me (though inevitably I'm going to fat-finger a push to origin)
>
> How are people supposed to read and comment on feature branches (I
> assume that's what you mean) of things that are proposd for merge to
> master?
>
> I would say it's fine for live feature branches to be in the main repo,
> but branches that are abandoned should be deleted (which is harder than
> it sounds, from experience).
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/postgis-devel



-- 
Raúl Marín Rodríguez
carto.com


More information about the postgis-devel mailing list