>> Maybe I'm some crazy, radical DBA, but I've never had a version of
>> pgsql get EOLed out from underneath me.
Just for fun, I did a bit of digging in the release notes
http://developer.postgresql.org/pgdocs/postgres/release.html
and came up with this table about PG major releases and their
follow-on bug fix/minor releases:
Version Release date # updates Days till final update Days till next major
6.0 1997-01-29 0 0 130
6.1 1997-06-08 1 44 116
6.2 1997-10-02 1 15 150
6.3 1998-03-01 2 37 243
6.4 1998-10-30 2 51 222
6.5 1999-06-09 3 126 334
7.0 2000-05-08 3 187 340
7.1 2001-04-13 3 124 297
7.2 2002-02-04 8 1190 296
7.3 2002-11-27 21 1867 355
7.4 2003-11-17 19+ ? 429
8.0 2005-01-19 15+ ? 293
8.1 2005-11-08 11+ ? 392
8.2 2006-12-05 7+ ? 426
8.3 2008-02-04 1+ ? ?
It's pretty clear that there was a sea-change around 7.2/7.3 ---
before that, nobody thought that PG releases were anything that
might be long-lived. And there's nothing in this table that
suggests we've really settled on a new lifespan ... other than that
we're still putting out new majors at a constant rate, and the community
hasn't got the resources or interest to maintain an ever-increasing
number of back branches.
regards, tom lane