On Fri, 2020-04-03 at 12:07 -0400, Ben Cotton wrote:
> On Fri, Apr 3, 2020 at 11:59 AM Leigh Griffin <firstname.lastname@example.org> wrote:
> > > Can we *please* see the final actual definitely official Fedora list,
> > > then? If this is supposed to be an open process?
> > @Ben Cotton can oblige here, it's not my place to share it without a stakeholder approval.
> The list sent to CPE is below. While there was no intent to hide it
> (it can be reconstructed from the council-discuss thread), it was a
> mistake on my part to not explicitly post this at the end of the
> As a Fedora contributor, I want the git forge to integrate with FAS so
> that it can use FAS to provide authentication and authorization.
> As a Fedora contributor, I want the git forge to integrate with
> fedora-messaging so that it can be a part of automatic workflows.
> As a Fedora contributor, I want it to be easy to add new contributors
> to a project (and optionally to enable self-adding) so that joining
> new teams is low-friction.
So, Leigh was correct,
Disclaimer that this is not aimed at you Adam, it's a broad statement. It's a shame that the perception is that I'm not correct or truthful on points like this. I can see why that might be the case given problems with the communication flow but know this, what I stated in all of my replies is truthful and up front about how we evaluated and discussed this. Whether we got it perfect or could have done things differently is another discussion.
and the F/OSS and self-hosting requirements are
entirely removed from this list. Not adjusted or de-emphasized or given
nuance, but simply removed entirely.
I highly dispute the idea that the removal of the F/OSS requirement
could be "reconstructed" from that initial list plus the discussion at
. I do not believe that is the case at all. Matthew's comment is
confusing and ambiguous, and there was no follow-up to it (at least,
not the F/OSS part of it), and it seems extremely questionable to me
that we would remove such a fundamental requirement based solely on one
confused comment from Matthew. He is the FPL, he is not the Pope.
The end result of this is that we (Fedora) have somehow indicated to
CPE that we have no preference whatsoever for F/OSS tooling. I do not
believe that should have been the case.
The self-hosting requirement at least Matthew was more clear in opining
should be removed, but it is still surprising to me that the process
here went "gather a list of requirements from the community, then if
Matt says he disagrees with one, take it out immediately but don't tell
anyone you did that"! (also I'll note that his substitute requirement
that out-migration be easy does not appear to be captured in the final
list, although it seems this probably *is* the case with Gitlab so we
don't have a problem there.)
One thing I'll note here: this is *exactly* the kind of thing that
would have come to light very quickly if the open process which was
committed to at the start had actually been followed through on.
It may have caught that for sure but it would have opened a lot more problems as stakeholders try to counter each others requirements with new more specific requirements to influence the decision. The approach taken, for better or worse evaluated it at a functional level without that noise factor. The trade off is losing an opportunity like this, however it could have been picked up several other ways and Ben has already apologised for not sharing back the final list.