killing thread ?

[Paul Rubin]

Peter Hansen <peter at engcorp.com> writes:
> > >   Whether a feature is supported *in the core language* and the
> > > *official distribution* -is- up to the language developers, though,
> > > especially in the case of free software.
> > 
> > Yes, indeed, the language developers are perfectly entitled to release
> > programs that don't live up to the marketing claims written by those
> > same developers.  However, nobody should then be surprised when users
> > complain that their expectations have been let down.
> 
> I agree.  Fortunately, as far as I can tell no one was surprised
> in this case when a user didn't have his expectations met.  I
> did see lots of explanations as to *why* he can't do what he wants,
> but very little suggesting surprise.
> 
> Anyway, the argument is kind of moot, isn't it?  Especially considering
> how often it has been raised, and how consistent the answer is to date.

There's an entry in Python's bug tracking system on Sourceforge about
supporting asynchronous exceptions in multi-threaded programs, showing
the developers consider it something worth looking for ways to
implement, rather than something that can't be done or something that
shouldn't be done for portability reasons.

Since the new "minimal Python" is supposed to be stackless (all
activation records are in the heap), maybe that implementation can
more gracefully deal with the problem of killing a thread failing to
unwind the C stack.