Re: Formatting Curmudgeons WAS: MMAP Buffers

Tom Lane wrote:
> Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> > On Mon, May 9, 2011 at 11:25 AM, Bruce Momjian <bruce(at)momjian(dot)us> wrote:
> >> Interesting. You could argue that once 8.3 is our earliest supported
> >> release that we could even shrink the support window because the
> >> argument "I can't dump/reload my data" would be gone.
>
> > Personally, I think the support window is on the borderline of being
> > too short already. There are several Linux distributions out there
> > that offer 5-year support for certain releases.
>
> Keep in mind that at least some contributors are paid to do exactly that
> long-term support (and if you've not heard, Red Hat is up to seven years
> support on RHEL ...). So the work is going to get done, and if it
> doesn't get committed to the community SCM, I'm not sure that really
> helps anybody.
>
> Although whether we do formal releases is a different question. Maybe
> it would be sensible to continue patching an old branch but not bother
> wrapping up release tarballs? But the incremental work to do one more
> set of release notes and one more tarball build is not that large.

I think the big reason we trimmed the support window was to push people
off of old releases, not to lighten our workload. Until we stated that
a release was not supported, we didn't give administrators ammunition to
force upgrades within their organizations.

Yeah, that is a lousy reason, but it was the stated case when we shrunk
to five years. You can argue that our more recent releases are not as
"stop using them" bad as previous ones, but Greg Smith's statement about
autovacuum badness reinforces that goal.

--
Bruce Momjian <bruce(at)momjian(dot)us> http://momjian.us
EnterpriseDB http://enterprisedb.com

+ It's impossible for everything to be true. +