Finding cycles and garbage, was( RE: Python Productivity over C++)

[Mike Fletcher]

Hmm, Neil's patch doesn't seem to be compatible with Stackless? (this app is
stackless-dependent (10s of 1000s of micro-threads of control, huge numbers
of waiting/suspended/manipulated micro-threads)).  Cyclops (in my fumbling
hands) hasn't come up with any notable cycles (there was a single one that
kept a single (small) object in memory, but that was the stackless
scheduler, which is seldom created/destroyed).  I'm looking for something
lower-level, like the VC++ feature where, at the end of the run, the
debugger says "hey, as I was doing shutdown, I noticed that the following
objects (including strings, floats, etceteras) were still in memory, that's
probably not what you wanted".  Neil's patch should be able to do that...
just need to get that and stackless into core Python and we're rocking :) .

Ah well, servers were meant to be restarted every 60,000 hits or so ;o) .

Enjoy,
Mike

-----Original Message-----
From: Michael Hudson [mailto:mwh21 at cam.ac.uk]
Sent: Friday, June 09, 2000 10:56 AM
To: python-list at python.org
Subject: Re: Python Productivity over C++
...
Have you tried Neil Schemenauer's patch?

http://www.enme.ucalgary.ca/~nascheme/python/gc.html

It can help with some of these things.

There's also cyclops, which I think is somewhere on ftp.python.org.
...