Re: Moving VAX into 21 century :-)

[Mouse <mouse%Rodents-Montreal.ORG@localhost>]

> [...] IMHO portability is the *only* point of HLLs [...]

Personally, I write in HLLs even when I'm writing code that's not
intended to be portable, simply because they improve my effectiveness:
I can, usually, get a program for a given task working sooner by
writing in an HLL than in even assembly, never mind machine code.

Consider the OCaml code used by the people behind FFTW to search for
FFT algorithms.  It's even turned up algorithms they "do not completely
understand".  Do you actually think it would be feasible for _anyone_
to implement that search in assembly?  I am inclined to doubt it; I'm
quite sure there are significantly fewer who could do it in assembly,
even if that number is nonzero, than who can do it in OCaml, even
taking into account the relative unpopularity of functional languages.

> (if you're going to write non-portable code, why let it be fat and
> slow too?).

(1) HLLs do not necessarily make for fat and slow code; for some
architectures, I've seen it said that compiler slop is actually
negative, and my own experience makes that plausible.  (2) As I
sketched above, I reject the implicit claim that anything a human
writes in a high-level language, a human could write in assembly (I
consider you to be making that claim because, without that, there's
nothing for the HLL version to be fat and slow with respect to).

> Also:  agree about not needing a FP, other than as a convenience for
> debugging.

I don't consider debugging a "convenience".  I consider it essential.
Perhaps this just means I'm incompetent.

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B