Re: gh_enter reconsidered

[Maciej Stachowiak <mstachow@mit.edu>]

jimb@red-bean.com writes:
> 
> > Frankly, though, Guile has a lot of problems right now, none of which
> > have anything to do with the GC it uses. There's no documentation, it
> > is slow, it doesn't have the sort of interfaces or module system it
> > needs, etc. It isn't a worthy competitor to things like Perl, at least 
> > not yet. Once the basic problems are further along, maybe spending
> > time on GC would be good....
> 
> "[Guile] is slow" has a lot to do with the GC it uses.  The
> documentation is more important, but improving the GC is still quite
> important.
> 
> I said that Guile should continue to use a conservative GC.  I don't
> at all mean to discourage people from improving the GC within that
> requirement.

I think GC is about third on the list of usual suspects, behind the
module system and insufficiently optimized primitives (regexp, file
I/O, etc). Improving GC performance has the additional advantage,
however, that GC can be run more often, reducing memory consumption.

 - Maciej