[FFmpeg-devel] [RFC] Experiment: enable github pull requests

[Kieran Kunhya]

On Wed, 12 Feb 2025, 23:29 Timo Rothenpieler, <timo at rothenpieler.org> wrote:

> On 12.02.2025 23:16, Soft Works wrote:
> >
> >
> >> -----Original Message-----
> >> From: ffmpeg-devel <ffmpeg-devel-bounces at ffmpeg.org> On Behalf Of Timo
> >> Rothenpieler
> >> Sent: Mittwoch, 12. Februar 2025 23:05
> >> To: ffmpeg-devel at ffmpeg.org
> >> Subject: Re: [FFmpeg-devel] [RFC] Experiment: enable github pull
> requests
> >>
> >> On 12.02.2025 23:01, Soft Works wrote:
> >>>
> >>>
> >>>> -----Original Message-----
> >>>> From: ffmpeg-devel <ffmpeg-devel-bounces at ffmpeg.org> On Behalf Of
> >> Timo
> >>>> Rothenpieler
> >>>> Sent: Mittwoch, 12. Februar 2025 22:51
> >>>> To: ffmpeg-devel at ffmpeg.org
> >>>> Subject: Re: [FFmpeg-devel] [RFC] Experiment: enable github pull
> requests
> >>>>
> >>>> On 12.02.2025 22:37, Soft Works wrote:
> >>>>>
> >>>>>
> >>>>>> -----Original Message-----
> >>>>>> From: ffmpeg-devel <ffmpeg-devel-bounces at ffmpeg.org> On Behalf Of
> >>>> Timo
> >>>>>> Rothenpieler
> >>>>>> Sent: Mittwoch, 12. Februar 2025 22:33
> >>>>>> To: ffmpeg-devel at ffmpeg.org
> >>>>>> Subject: Re: [FFmpeg-devel] [RFC] Experiment: enable github pull
> >> requests
> >>>>>>
> >>>>>> On 12.02.2025 22:22, Stephen Hutchinson wrote:
> >>>>>>> On 2/12/25 2:23 PM, Lynne wrote:
> >>>>>>>>
> >>>>>>>>
> >>>>>>>> On 12/02/2025 19:53, Romain Beauxis wrote:
> >>>>>>>>> Le mer. 12 févr. 2025 à 12:49, Lynne <dev at lynne.ee> a écrit :
> >>>>>>>>>>
> >>>>>>>>>> Users can already login to the instance with their github
> accounts, so
> >>>>>>>>>> there's no barrier to entry.
> >>>>>>>>>
> >>>>>>>>> Would you mind sharing the url? A quick search didn't reveal
> >> anything.
> >>>>>>>>
> >>>>>>>> Sure
> >>>>>>>> https://code.ffmpeg.org/
> >>>>>>>>
> >>>>>>>
> >>>>>>> Are all accounts restricted to owning a maximum of 0 repositories
> by
> >>>>>>> default, or is it set to 0 only for those that sign up through one
> of
> >>>>>>> the external logins?
> >>>>>>
> >>>>>> It's set to 0 by default, to avoid spammers uploading junk, or just
> >>>>>> people (ab)using it for non-ffmpeg things.
> >>>>>>
> >>>>>> You can open issues and comment on existing PRs.
> >>>>>> And also create PRs using the AGit workflow:
> >>>>>> https://forgejo.org/docs/latest/user/agit-support/
> >>>>>>
> >>>>>> The repo limit can manually be lifted per user. I preferred that
> >>>>>> approach vs. the Videolan approach of completely locking down the
> >>>>>> instance, and requiring admin approval for every single new user,
> which
> >>>>>> imo is more detrimental to new contributors than a ML.
> >>>>>
> >>>>> Hi Timo,
> >>>>>
> >>>>> Does that mean that you cannot create forks and create PRs from that
> >> forked
> >>>> repo like on GitHub?
> >>>>>
> >>>> Not sure what you mean, you need "admin approval" to be allowed to
> >>>> create repos, including forks.
> >>>
> >>> I don't think it's a good idea to build such entry bars.
> >>>
> >>>> Just to avoid abuse. Obviously you can then PR from that fork once you
> >>>> got one.
> >>>> Or you can submit PRs without a fork as stated above.
> >>>
> >>> That's a workflow I've never heard about - same like most other
> >> contributors. Again, that is building an entry bar.
> >>>
> >>> I think people should be able to use a procedure they are familiar
> with.
> >>> Is it possible to create PRs from a fork on GitHub?
> >>>
> >> I'm really not sure what you're asking.
> >> PRs are not restricted. Creating repos is.
> >> And there is no way to NOT restrict it, unless you want to pay several
> >> hundred Euros a month in hosting fees extra, and constantly be on the
> >> lookout for hosting illegal/harmful things.
> >
> > I wasn't asking, I'm stating that not being able to use an established
> workflow like
> >
> > fork >> clone >> develop >> push >> PR
> >
> > ...would be an entry-bar for new contributors.
> >
> > But here comes a question: I've read that the "AGit flow" work by
> creating a branch for each submission in the original repo. Doesn't the
> repo get "polluted" over time this way? In case of merged PRs, the branch
> might get deleted, but what about unmerged ones?
> > And when one clones the whole repo, don't they get all those branches
> downloaded locally as well? (as long as one doesn’t specify which branches
> to download)
>
> So do forks, cause they all live in the same underlying repo as well.
> Unless a user deltes the repo/branch/PR, stuff pushed there will hang
> around forever.
>

Yes but surely if users do a simple clone of the main repository, they will
get all the branches of hundred of different pull requests?

Kieran

>