psyco, jython, and python speed tests

[Sandy Norton]

Kragen Sitaker <kragen at pobox.com> wrote in message news:<83hendkou7.fsf at panacea.canonical.org>...
> sandskyfly at hotmail.com (Sandy Norton) writes:
> > o Developments like psyco should be encouraged and supported because
> >   ultimately accelerating python is a good thing: it takes away one 
> >   usual criticism of those who have _never tried_ to develop in 
> >   python. For those who have, the productivity benefits, of course,
> >   by far outweigh any so-called 'performance' issues.
> 
> This is unjustifiably glib.  Lots of programs don't need to run fast.
> Some programs do need to run fast.  Some programs justify significant
> work to get them to run fast.  Accelerating Python is a good thing not
> because of "so-called 'performance issues", but because in some cases,
> there really *are* performance issues that rule Python out.

I don't disagree with you except on the glibness part (-; Perhaps if I
elaborate on the unapparent context of this remark you may see were it
came from: a colleague of mine started raving about python recently,
and now wants to incorporate the jython implementation into his
developments efforts. He explained that he had resisted using python
because he had 'heard somewhere' that it was slow... After actually
giving the language a shot he changed his mind.

I'm not denying that pure python may not be appropriate for certain
applications which require maximum performance, but that's why you
have numerous ways to glue optimized parts into the nicely high-level
whole.

python wouldn't be python without this marvelous extensibility...

regards,

Sandy