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

Re: when GC is permitted

This page is part of the web mail archives of SRFI 50 from before July 7th, 2015. The new archives for SRFI 50 contain all messages, not just those from before July 7th, 2015.



Tom Lord <lord@xxxxxxx> writes:
> It's a clearer explanation.  I'm not sure how useful it is to solve
> the problem.  Consider for example that, applying the rule as you've
> stated it, all calls to any function that accepts or returns a Scheme
> value must be made from within a gc-excluding begin/end pair.

Both you and Felix have misunderstood where I was coming from.  I
don't support Kelsey's proposed solution as adequately addressing all
the concerns we've raised about SRFI-50.  I still don't like it.  But
within the priorities he's explained, it does address the problem I
raised.  Everything I wrote was simply a response to your assertion
that Kelsey had said something nonsensical --- not an indication that
I think SRFI-50 is peachy as amended.

Similarly with my reply to Felix.  I think the overhead of the Minor
API vs. SRFI-50 is acceptable.  For heavy Scheme object munging, if
you've got a good compiler, you should write Scheme code.  But two
orders of magnitude *is* a significant amount of overhead, and Kelsey
isn't some kind of dope for being concerned about it, given that he
can't escape to Scheme code for speed.