forge vs gitolite (Re: CTI - Making a decision for glibc.)

[Florian Weimer]

* Alexandre Oliva via Overseers:

> On Jan 31, 2026, Simon Marchi <simark@simark.ca> wrote:
>
>> On 2026-01-31 02:11, Alexandre Oliva via Overseers wrote:
>>> Do you have any plans to avoid the SaaSS features of typical forges?
>
>> Can you explain what the SaaSS features of typical forges are?
>
> I don't have an exhaustive list, but the computing activities that first
> come to mind, that I've already mentioned in this thread, are patch
> merging and CI/CD.

I don't understand this kind of objection.

It's pretty much a given that in a functioning developer community that
people run software on their machines on your behalf, even if you could
in theory run the testing locally on your own hardware (if you owned
such hardware).  CI/CD merely automates that process.  Of course it
cedes control, but without that, collaboration isn't possible.

Keep in mind that they I don't think there should be a firm line between
developers and users.  I believe software freedom requires that users
can turn into developers.  Forges and CI/CD can be an important tool to
enabled.  I'm rather sad that the users-to-developers idea is somewhat
controversial within the GNU project (and my impression is not mainly
based on opposition to forges, it's that many components are so very
difficult to build from source on common platforms).

Thanks,
Florian