Re: Release planning

Christopher Browne wrote:
> A long time ago, in a galaxy far, far away, Gaetano Mendola <mendola@bigfoot.com> wrote:>>>>I was thinking of
somethingmuch simpler where Jan would create an>>>ARC patch against 7.4.X and have it either in /contrib for 7.4.X
or>>>onour ftp servers, or on a web site.  I could create a mechanism>>>so SELECT version() would display Jan's
add-on.>>>>:-(>>>>Iwas asking to add the vacuum delayed patch to 7.4 months ago and>>the response was: why introduce
instabilityto a stable release ?  I>>hope the global consensus is a no way to procede also for ARC.>>> If, as you
suggest,ARC is too immature to go in, then I presume that> would also imply that "global consensus" should also be
that:>> a) PITR is much too immature to put in;>  b) Win32 is way too immature to put in;>  c) NT is way too immature,
aswell;>  d) Tablespaces are way too immature to be included.>> Which eliminates all of the big, interesting reasons to
upgradeto> 7.5.>> Or is ARC the only thing you regard as a misfeature?  Interesting that> it was put into 7.5 _early_
inthe process, so that everyone that has> lately touched the betas would be getting bitten pretty heavily if it> was
ladenwith bugs...
 

No no, please don't mistake me. I wrote that shall be a no way to back patch
ARC in the 7.4 version, starting from consideration that the delayed vacuum
was considered too dangerous...

My humble opinion is that have all that new funcionts in one shot with the same
ammount of beta period for other release is too dangerous. From my side the
only think that I can do with a beta is just run our regression tests ( almost
2000 tests ) for our installation, but for sure I can not use it extensively.
We could ( may be is a crazy idea )release the

7.5 with ARC + BW + new list implemetation
7.6 with 7.5 + PITR
7.7 with 7.6 + NT
7.8 with 7.7 + Win32
7.9 with 7.8 + Table Space

with 2 months between a release and the other with a one month of beta for
each release; this in order to reach the version 8.0 rock stable; and possibly
without need to initdb between these releases.

Is it too insane ? :-)

I'm sorry to see Postgresql releases driven by advertisment instead by good sense
( as it was till today ).



Regards
Gaetano Mendola