[Stackless] Compatibility?

Richard Tew richard.m.tew at gmail.com
Wed Aug 9 12:32:53 CEST 2006

On 8/9/06, Ethan Glasser-Camp <glasse at cs.rpi.edu> wrote:
> Richard Tew wrote:
> > Yes, there are problems with C extensions.  But most C extensions
> > I used are pretty straightforward and I have not experienced the
> > problems myself.  I think it is to do with the stack switching and
> > these extensions not being compatible with it and crashing
> > because of it.  You should be able to find more relevant posts about
> > these problems on the mailing list archives.
> Thanks for the pointer. I think the thread you mean is:
> http://www.stackless.com/pipermail/stackless/2003-February/000655.html
> which says that "if you switch from an extension which relies on
> variables being alive on the stack, you got a problem". But in order
> for this to happen, C extension code has to call Python code that
> switches tasklets to something that came in from outside the extension
> code, so it is unlikely to cause compatibility issues.

That sounds right.  I didn't mean any particular thread, but I expected
Christian would have stated the problem at some point.  Thanks for
searching it out.

> I also see this thread from a few months ago:
> http://www.stackless.com/pipermail/stackless/2006-May/001741.html
> which seems to suggest that Stackless aims to be binary compatible,
> and if it's not, people would like to make it be. Is that so?

Well, in the sense that bugs need to be fixed, yes.

But there is no reason I know of that PyQT should not be binary
compatible, and I personally have little interest in using it, so there
would be no rush on my part to look into it.  I might install PyQT
and import it and debug that, which shouldn't take much of an
investment of time and hopefully the problem will be PyQT doing
something horrendous so I can claim it is their problem :)

I get the feeling you are leading up to something with this question?
Are you going to offer to look into this problem? :)


Stackless mailing list
Stackless at stackless.com

More information about the Stackless mailing list