So, some general thoughts here…
Thanks for starting off this discussion. I fear it’s going to be a long
one, but it’s one we need to have. 
I agree that using a general ‘forge’ isn’t ideal for packaging, but I am
not sure I understand where any development of a net new infrastructure
just for that would come from and would mean we would be mostly ‘going
it alone’ like we are for pagure right now. There is a lot of overlap
between what package maintaing needs and what a forge provides. Using a
more generic solution allows us to share maint with everyone working on
that forge too.
I mean, it’s all back to the ‘off the shelf’ vs ‘custom’ software
discussion thats been around forever.
I like the idea of a small list of “MUST” and larger lists of “NICE”,
but I also agree that spending lots of time gathering requirements may
not be the right way to go, but I think we either need something or a
good feedback loop to tell people what are hard requirements.
There’s a swirl of some other things that all surround this too:
- As mentioned already, bugs and moving from bugzilla.
- source git? There was a sig, but it ran out of time/people, but this
would be a time to examine if a ‘new’ thing would make more sense
there.
Perhaps we could break this into proposals and have people who are
passionate about each one put together a proposal and then we can crowd
source adding pros and cons or providing discussion. Or perhaps we could
come up with an initial list of ‘possible’ and do proposals after that?
-
Keep using pagure. What would this look like? what do we consider
‘healthy’ enough? -
Write our own package maintaining application. Who would do this?
what would it look like? could we maintain it long term? -
gitlab ( https://about.gitlab.com/ )
-
Forgejo ( https://forgejo.org/ )
-
Radicle ( https://radicle.xyz/ )
-
Fossil ( Fossil: Home )
-
sourcehut ( https://sourcehut.org/ )
-
allura ( https://allura.apache.org/ )
-
…your idea here…