New Python User Question about Python.

Tim Peters tim.one at home.com
Fri Aug 24 01:50:28 EDT 2001


[Chris Barker]
> ...
> Apparently, a major effort to optimise for speed is not on the top
> of the developers list,

That's true, but it appears to be mixed with a misunderstanding of reality:
there's no major effort of *any* kind on the developers' list, neither at
the top nor at the bottom.  120% of available time is consumed by an endless
sequence of small but achievable projects; the biggest thing going today is
the type/class unification, and the total effort for that is hoped to be
just a few person-months.

Earlier this year I spent two intense weeks *just* trimming runtime fat out
of Python's dict implementation.  Do you have an idea what it takes to
undertake a comprehensive optimization project?  I'll immodestly claim that
I do, and I see no way for PythonLabs to get there from here.  The best hope
you've got for major Python speedups is the vague hope that Python can
piggyback on some other group's work.  Short of that, you'll see a sequence
of small but achievable improvements.  In other words, the next 10 years
will probably look like the last 10, surprisingly similar to real life in
that respect <wink>.

although-i'll-get-even-grumpier-ly y'rs  - tim





More information about the Python-list mailing list