On 15/11/2023 04:50, Mike Rochefort via CentOS-devel wrote:
The potential impact when we'll just decommission mirrorlist.centos.org is for SIGs building for RHEL (and not Stream) 8 and so still relying on mirrorlist= in their shipped .repo. The list of impacted SIGs should be really small and as the content itself will probably moved to mirror.stream.centos.org, there will be a different thread to discuss how to do that
I've been wondering what the plan for CentOS Stream content has been for a while. Though since CentOS Stream 8 will EOL before CentOS Linux 7, I guess it's a moot point and only EL8 SIGs supporting RHEL will be relevant to move off of m.c.o.
That was the plan yes, ensuring that SIG content built against/for RHEL8 would move to mirror.stream.centos.org, the same way that it's done for RHEL9 : see the 9-stream vs 9 layout under https://mirror.stream.centos.org/SIGs/ , so we'd have then 8 landing there too
# CentOS Forums (https://forums.centos.org) We discussed this already in the past (on this list and elsewhere) : while we can quickly migrate that service to newer stack, the underlying question was more about current moderators saying that nobody will maintain that service after June 2024. So in the current situation, we'll just decommission the forums in 2024 with no replacement. Should the situation change between now and June 2024, we'd be able to migrate (or not) the service.
Should we be taking this opportunity to promote and direct CentOS Stream users to the CentOS category on the Fedora Discussion[0] platform? Similar to the Asahi Linux category, we could pitch up sub-topics to filter Project discussion and community support.
AFAIR there is (was ?) already a centos category on discourse but don't know if it's active or not
While instant messaging and email platforms are available, I believe having an easier point of entry for longer lived and easily searchable material is a benefit to the community at large. Letting it passively wither or just outright decommissioning the platform without a similar service is not an ideal situation. Instant messaging is easy to get lost in and lose track of important discussions even with available *in-platform* search, and email is... well I don't think the vast majority of people on the net care for it that much or know how to use it "correctly". Maybe migrating to a HyperKitty platform like lists.fp.o[1] could help with that?
Yes, and I also mentioned mailman3 migration for centos in this thread ;-)
# centos linux 7 buiders (infra) Not publicly visible but pretty obvious that we'll just keep that infra running until CentOS Linux 7 EOL and then we'll just decommission it entirely (as we'll not have to care about building/shipping anything for centos linux 7
If there are hardware resources involved, would they be wholly decommissioned or repurposed for CentOS Stream + SIG usage?
All the hardware used to build centos linux 7 artifacts is already out of warranty and lower specs than the SIGs builders so I think that hardware will just be decommissioned completely