On Thu, Jun 10, 2010 at 11:30 PM, Simon Riggs <si...@2ndquadrant.com> wrote: > On Thu, 2010-06-10 at 22:29 +0800, Russell Keith-Magee wrote: > >> There appears to be some confusion here. We're not recommending a >> version of PostgreSQL that end-users should use; we're nominating the >> minimum possible version that passes Django's test suite. PostgreSQL >> 7.4 may be bad for all sorts of reasons, but right now, all the >> features of PostgreSQL that Django uses utilize are available in >> PostgreSQL 7.4. When the fix for #8901 lands, this raises the bar to >> PostgreSQL 8.0, but only because of the use of >> pg_get_serial_sequence(). >> >> I completely agree that users would be well advised to upgrade to >> PostreSQL 8.4, and there are many reasons beyond basic Django >> compatibility that should drive that upgrading process. However, that >> doesn't change the fact that from a purely functional perspective, >> Django will work happily on PostgreSQL 8.0 (unless you want to use >> database-level autocommit or savepoints, in which case the minimum >> version is 8.2). > > I take your point about the differences. > > In this case, the PostgreSQL project is just about to de-support those > release levels, so that does change things somewhat. I'm not sure why it > would be useful for a future version of Django to advertise support for > a PostgreSQL version that the PostgreSQL project is itself intending to > de-support. That seems likely to cause disappointment, even though you > are correct and it will pass tests.
I suppose that depends on how you interpret "advertise". Keep in mind the reason that this issue arose in the first place -- a user had a Django install on a VPS that provided PostgreSQL 7.4, and got bitten by a changeset that inadvertently changed the minimum supported version. I'm in complete agreement that this user *should* upgrade to a newer version of PostgreSQL, but that doesn't change the fact that Django 1.2 *did* work on PostgreSQL 7.4. If PostgreSQL 8.0 is still officially supported by PostgreSQL, and it works fine under Django, then there's no reason to take it off the supported versions list. The only real reason to take something off *our* supported version list is when supporting that version imposes an engineering overhead that we're not willing to bear. I'm certainly not going to put any engineering effort into supporting a version of PostgreSQL that PostgreSQL itself isn't trying to maintain, but until we actually need to make that call, I'm happy to "support" 8.0. That doesn't mean we can't encourage people to use newer versions, though. The MySQL docs already contain hints in this direction, saying that MySQL 3.23 will probably work, but you'll have less trouble if you use 4.1 or 5.0. I suspect a similar hint may be in order for PostgreSQL. By the time 1.3 comes out (ETA approx Dec 2010), 8.0 and 8.1 will both be EOL, so it may be prudent to include a note to the effect that 8.0/8.1 will "work", but they're not supported by PostgreSQL anymore, and there are lots of advantages outside basic Django compatibility to using 8.2+ that are well worth the effort of upgrading. > I am just the messenger in this, carrying goodwill between projects. Most appreciated. Yours, Russ Magee %-) -- You received this message because you are subscribed to the Google Groups "Django developers" group. To post to this group, send email to django-develop...@googlegroups.com. To unsubscribe from this group, send email to django-developers+unsubscr...@googlegroups.com. For more options, visit this group at http://groups.google.com/group/django-developers?hl=en.
