Re: 'Official' definition of ACID compliance?

On Thu, 2006-01-05 at 11:39, John Dean wrote:
> At 16:38 05/01/2006, Stephen Frost wrote:
> >* Russ Brown (pickscrape(at)gmail(dot)com) wrote:
> > > Oh, that's a long story. We're a MySQL house that I've been trying to
> > > convert to PostgreSQL one way or the other for ages (with no success as
> > > yet). Note that the argument isn't about which letter the type
> > > truncation applies to, but whether it actually has anything to do
> > > with ACID at all in the first place. The key for me is that the result
> > of this argument has an
> > > effect on the question: "Is MySQL ACID compliant". If I'm right, it's
> > > not (which has political strategic benefits to me).
> >
> >An even better thing to point out is that a DBA recommending MySQL isn't
> >a DBA at all. :)
> >
> > Enjoy,
> >
> > Stephen
>
> I used to work for MySQL (a job's a job after all) and I say in all honesty
> that MySQL is not ACID compliant. Furthermore, MySQL is so lacked in
> functionality that it should be used for anything but the simplest of
> solutions. A database engine that does not support referential integrity,
> triggers, stored procedures, user defined types, etc should not be taken
> seriously

PHP 5.0 has most of those features now. It's just the inability of the
DBA to force things like certain tables to be used that I hate about it.

That and even in V 5 it sill ignores row level foreign key definitions
(they have to be done at the end of the column list) silently.

I bet in another year or two MySQL will be breathing down the neck of
PostgreSQL V 6.5.3 in terms of features and proper operation.