XCP is useful and works well for some people - is there any reason not
to simply leave the last released version in the archive and stop trying
to package new versions? Even if those don't have the latest features,
they are still useful in comparison to alternatives.
There are many products (e.g. procmail) that are extremely popular but
have not been formally supported by any upstream for years[1]. Mozilla
publicly announced an end to development of Thunderbird[2], but it
remains popular too.
Maybe we need to have a mechanism to let people see the upstream support
status of packages at time of installation and make their own decision
if they feel comfortable with it?
Could anybody add any comments to this bug about migration strategy?
Should people anticipate using CentOS for dom0 systems in future? Or
just stick with Xen-API on wheezy (which will still be around and
subject to security updates through to 2016)? Or move to KVM or Ganeti
or something else? It would be interesting to see a comparison of the
options, and the impact on things like OpenStack packaging, even if
there is no clear solution for all users.
It sounds like Xen-API is at risk of ending up like FreeSWITCH and other
products that are unpackagable (a common problem in the VoIP space).
While some of them to remain 100% free (just inconvenient, due to
bundled libraries that upstream forked) there is always a risk that as
things move out of the mainstream, they do not remain free software at all.
Big thanks to the maintainers for all the work they have put in up to
this point.
1. http://www.procmail.org/procmail.HISTORY.html
2.
http://www.zdnet.com/mozilla-scraps-thunderbird-development-email-client-not-a-priority-anymore-7000000469/
