Stackless/microthreads merge news

Glyph Lefkowitz glyph at
Tue May 16 18:09:57 EDT 2000

Okay.  I wanted to make a point, that C. Tismer's time would be better
spent with something other than making a stackless Java VM ... and
that point having been made, we should leave this little subthread...

ObOnTopic -> One thing that's been bugging me lately though; Java is
definitely faster than python, but Python and Java both use virtual
machines.  The difference, as far as I can tell, is that Python
objects (namespaces, classes, class instances, etc) are actually all
*hashtables*.  Java's are similiar to C structs, at least in how bytes
are arranged, linked, and accessed.  Java's speed advantage is
significant, but from what I've seen so far, it doesn't seem to be the
massive leap beyond python that I'd expect.

Is there a good technical reason for this, or is it just a poor
implementation?  I haven't had a chance to read enough of Python's
sources yet, I guess, but is it doing any tricks with attribute access
to make them faster than regular hashtables?  (or is my estimation of
the speed difference really just erroneous?)

The obvious next question is "if not, why not?" but I think I
understand that, if it's the case ^_^.

JRHoldem <jraddockNOjrSPAM at> writes:

> I'm new to this board but you're right - Glyph is a programming
> god and he's 100% on target with his Java rants. I say this not
> only as a business partner and friend of his but also as a Java
> programmer who has faced several of these issues myself.

This is so blatant, I almost want to accuse myself of using tentacles
;-) (thank you, JR)

> And yes, any language which has attracted the attention of the
> Cobol community probably is the work of Satan.

Iä!  Shub Microsystems!

        -I-should-be-more-on-topic-ly y'rs,

                 |    ______      __   __  _____  _     _   |
                 |   |  ____ |      \_/   |_____] |_____|   |
                 |   |_____| |_____  |    |       |     |   |
                 |   @ t w i s t e d m a t r i x  . c o m   |
                 |   |

More information about the Python-list mailing list