Re: single or double cell?

[Michael Livshin]

Dirk Herrmann <address@hidden> writes:

> [...]
> what I rather meant is whether it makes sense to use this
> knowledge outside of the gc and the corresponding types.  In the last year
> or so, a lot of work has been put into guile in order to break up
> dependencies between parts of guile, for example the dependency between
> the gc and the type system.  Thus, I am very skeptical towards approaches
> that introduce such couplings.

while you are absolutely right in general, I'd argue that in this
particular case (binary dumping/undumping) the closer-than-comfortable
relation with the allocator is basically inevitable.

perhaps some more-low-level-but-still-public interface to the
allocator could be defined to allow some kind of reflection over
object representation.  which reminds me that I even worked on
something related a year ago.  hmm.

-- 
The only thing better than TV with the sound off is Radio with the sound
off.
                -- Dave Moon