Aaron W. Hsu scripsit:
> I strongly suspect, but have no proof, that the cost of having a safe
> write procedure is dwarfed by most other costs, and will not contribute
> significantly to the running time of the vast majority of programs, if
> it contributes significantly to any at all.
[...]
> I think we are past this. We should not concern ourselves with hacks
> to improve the performance of implementations that do not care about
> performance. If performance is important, then use one of the already
> useful and good Scheme implementations out there that produce fast
> code. If you do not care about speed, feel free to use a naive
> implementation, but do not expect it to be fast.
You utterly ignore the question of space. Write/fast-small is not just
write/fast, it's also write/small: it can traverse trees requiring stack
space proportional to the depth of the tree and with no auxiliary space
at all. Write/safe and write/cycle *cannot* do so.
There is, and I believe always will be, a need for implementations
in space-constrained environments. If phones are no longer
space-constrained, there are still "dust", "skin", and "clay" (see the
Wikipedia article on ubiquitous computing). For those things, space
beats speed handily, especially when executing embarrassingly parallel
algorithms (which is what the Real World has to do).
> Again, if you care about performance at all, then you should not use
> an implementation that does not care about performance. It will not scale.
That sounds tautological, but what you are saying is "If you care
about speed at all, you should not use an implementation that is not
designed for maximum speed." That is not only not a tautology, it's
downright false. Bytecode interpreters will never be *fast*, but they
are faster than tree-walking interpreters, faster enough to make them
well worth writing and using in environments where optimizing compilers
are emphatically the Wrong Thing.
--
weirdo: When is R7RS coming out? John Cowan
Riastradh: As soon as the top is a beautiful co...@ccil.org
golden brown and if you stick a toothpick
in it, the toothpick comes out dry. www.ccil.org/~cowan
_______________________________________________
r6rs-discuss mailing list
r6rs-discuss@lists.r6rs.org
http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
