Slow tests

[Martin (gzlist)]

On 18/09/2009, Martin Pool <mbp at canonical.com> wrote:
>
> Getting good coverage on python2.4 is desirable because our pqm runs
> that.  As long as we support using 2.4 it's probably good that pqm
> should run it because nobody else does.

Hey, *I* do, though might be the only person on the list who still
uses 2.4 primarily. Just don't think coding style tests are that
important for coverage.

> If they're compatible I
> suppose there's no harm using a faster one in 2.5+.

They are not, it is possible to do more complicated things at the AST
level than with the parser or dumb regexps. Whether anyone wants
automated tests for this kind of stuff rather than just letting code
review catch it is another matter.

Martin