guile-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Release 1.6 critical TODO items.


From: Marius Vollmer
Subject: Re: Release 1.6 critical TODO items.
Date: 11 Mar 2002 17:42:40 +0100
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.1

Rob Browning <address@hidden> writes:

> Marius Vollmer <address@hidden> writes:
> 
> > Rob Browning <address@hidden> writes:
> >
> >> Marius Vollmer <address@hidden> writes:
> >> 
> >> > gentemp created symbols that were unique in a given obarray, which had
> >> > not much to do with the global hash table that holds all (interned)
> >> > symbols.  when you had no other use for the obarray, using gentemp
> >> > with a specific obarray was equivalent to using a specific prefix.
> >> 
> >> OK, so do I have to worry about thread safety?
> >
> > Yes, I'd say so.  We should probably offer a syntactic form for
> > wrapping code that needs to run isolated.  Is there some SRFI about
> > this?
> 
> Hmm -- could you elaborate?  I'm not sure exactly what you're asking?
> I was just asking if I should wrap my syncase.scm gensym internals
> with a (lock-mutex) (unlock-mutex) pair...

Yes, that's what I wanted you to do.  I also speculated about how to
make this easier with some compact syntax like

    (begin-synchronized
      ...)

where the 'begin-synchronized' form is only being executed by at most
one thread at any one time.  This should encourage people to actually
synchronize their code properly.  Since this should be a common thing
to do, I was asking whether one of the threading SRFIs deals with
this.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]