Re: c++ -> guile threading

[Peter Amstutz <tetron@student.umass.edu>]

On Wed, 12 May 1999, Greg Harvey wrote:

> > I was thinking that the current module would be entirely determined by
> > the lexical context of the currently executing code, without regard to
> > which thread might be executing that code.  When I try to think about
> > a current module determined at run-time when there is more than one
> > thing running, my mind gets bent.
> 
> That's pretty much what I was thinking, though I didn't put it quite
> right ;) The problem is that the module operations (define-module, for
> example) operate on variables global to everyone, so if you do
> something like
> 
> (make-thread (lambda () (define-module (foo)) (display "bar"))), 
> 
> the module of every thread becomes foo (though eval-in-module will
> prevent that (oops ;)). 

I don't suppose anyone is working on taking all the guile global variablse
and sticking them in a structure so you could have a number of states
cooexisting in the system?  Why wasn't guile designed this way initially?

           ------------------ Peter Amstutz --------------------
           -------------- tetron@student.umass.edu -------------
           ------- http://www-unix.oit.umass.edu/~tetron -------
           -----------------------------------------------------