MySQLdb windows binaries for Python 2.5?? Yes, but from a World of Warcraft guild.

Jorge Vargas jorge.vargas at
Fri Jan 5 22:06:18 CET 2007

On 14 Dec 2006 16:59:15 -0800, Fuzzyman <fuzzyman at> wrote:
> johnf wrote:
> > >>
> > >> If you search the Help Forum of the MySQLdb project on SourceForge, you
> > >> will find a couple of people who have successfully built MySQLdb on
> > >> Windows for 2.5, and are willing to share their installers.
> > >> That's how I got my binaries.
well it seems noone has talk to Andy directly I build the one for
1.2.1_p2 for python2.4 ones it was done all I had to do is email Andy
and he posted it without problems

the biggest issue here is that building that driver in windows is a
pain, you need mysql headers, python headers and microsoft compilers a
total of around 2Gb, plus some bug I found that ended up being an
error in between the file and windows commandline.

also windows lacks the mysql-config tool which means you need to find
out a lot of flags which on *NIX are fill in automagically.

ahh did I mention you need to patch distutils to work with the "free
windows" compiler because the official windows python builds agains
the non-free MS compiler.

> > > This, apparently, is the extent of current Python support for MySQL.
> > > Want to install that executable, as root, on your production machines?
> > >
> > > This is embarassing for the Python community.  Perl and PHP come
> > > with MySQL support built in.  Python is out to lunch on this.
> > >
> > I couldn't disagree more.  That fact that no Database drivers are built-in
> > makes Python stronger - allowing Python to access any Data Engine that
> > supports DBI 2.0.  Of course I'm not including pickle in my assessment.
> >
actually in this case I disagree having this driver as buildin will
take much of the pain out of any user who tries to compile this
driver, it will also help in standarizing the drivers for example in
postgre there is more then one, I think we will be better off with
just one driver, that is one DBI2.0 driver, there are some experiments
outthere which make it more "pythonic" which shouldn't be excluded.

More information about the Python-list mailing list