On Mon, Feb 18, 2002 at 10:40:59AM +0100, Jean-Michel POURE wrote:
> Le Lundi 18 Février 2002 06:18, Andrew Sullivan a écrit :
> > and MySQL+InnoDB. It seems that such would be an apples::apples (ok,
> > maybe one's an overbred hybrid apple, but still an apple) comparison
> > at last.
>
> Dear Andrew,
>
> The main difference is that PostgreSQL allows server-side programming and
> does things smartly:
Sorry, I wasn't precise enough. I know that MySQL does not have all
the features of Postgres. I simply meant that, for ages, there have
been all sorts of benchmarks floating around purporting to show that
MySQL is faster than Postgres. None of these have ever taken into
account the transaction overhead that Postgres automatically incurs.
With InnoDB transactions, MySQL seems to have the same sort of
transactional overhead. So, a real comparison can be made.
Yes, all sort of other nice features will be missing. No-one needs
to convince me of the benefits of Postgres. But if MySQL _really_
has transactions now, and they really scale and all that, then we may
have something to learn from it (well, ok, others may have something
to learn from it. I can barely write 'Hello World' in C, so I'm not
going to be much use as a code contributor). And since InnoDB and
MySQL are both GPL'd, one would be able to examine the source for
bright ideas, if it turns out that MySQL+InnoDB is blazing fast with
(say) 50 concurrent users.
Anyway, it was just a suggestion. Sorry I brought it up, since it
has the potential to spark another holy war.
A
--
----
Andrew Sullivan 87 Mowat Avenue
Liberty RMS Toronto, Ontario Canada
<andrew@libertyrms.info> M6K 3E3
+1 416 646 3304 x110