Using Python for business app development

Mike Meyer mwm at mired.org
Mon Dec 30 03:12:05 EST 2002


Paul Rubin <phr-n2002b at NOSPAMnightsong.com> writes:
> Mike Meyer <mwm at mired.org> writes:
> > > I'm also put off MySQL by its developers' and advocates' past confusion
> > > (to say it charitably) regarding its limitations.  For instance, back
> > > when MySQL did not support transactions, its developers told prospective
> > > users that most everything that transactions get you can be done in the
> > > application.
> > The *developers* actually said this? That's scary, all by
> > itself. After all, the reason I'm using a database in the first place
> > is so I don't have to rely on the application developers getting the
> > data locking right. Rollback and the like are nice, but if the
> > database can't insure my datas integrity in the face of simultaneous
> > updates, then I've chosen the wrong tool for the job.
> MySQL 4.x supports transactions.  Previous versions had enough locking
> that you wouldn't corrupt your data with simultaneous updates, but
> performance could take a dive.

I'm aware that current versions of MySQL have transactions. I just
haven't found anything wrong with Postgres to make me want to examine
other options.

As for not being able to corrupt the data, I'm talking about integrity
constraints that can be violated by a single SQL statement, but will
be satisified by a correctly written transaction. As an application
developer, I have to get the transaction correct, but at least I don't
have to worry about somebody else finding the database in a broken
state between SQL statements. When I found myself considering lock
files to keep a MySQL database from losing that integrity was when I
went looking for a better solution.

        <mike
-- 
Mike Meyer <mwm at mired.org>			http://www.mired.org/home/mwm/
Independent WWW/Perforce/FreeBSD/Unix consultant, email for more information.



More information about the Python-list mailing list