Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!samsung!rex!lang
From: lang@rex.cs.tulane.edu (Raymond Lang)
Newsgroups: comp.sys.apple2
Subject: Re: Re- HLLs vs. Assembly
Message-ID: <7007@rex.cs.tulane.edu>
Date: 13 Apr 91 00:27:59 GMT
References: <9104060651.AA18946@apple.com> <1991Apr6.100927.21953@nntp-server.caltech.edu> <1991Apr9.150402.563@latcs2.lat.oz.au> <15782@smoke.brl.mil> <6965@rex.cs.tulane.edu> <1991Apr10.203703.21010@nntp-server.caltech.edu>
Organization: C.S. Dept, Tulane Univ., New Orleans, LA
Lines: 50

In <1991Apr10.203703.21010@nntp-server.caltech.edu> toddpw@nntp-server.caltech.edu (Todd P. Whitesel) writes:
>Yes, but modifying the program to improve performance requires

>       (a) better overall algorithms &/or data structures
>       (b) knowledge about the hardware and what the compiler will produce

I don't follow how (a) comes from experience in assembly language.


>[convex example deleted]

>The point of all this is that while HLL's insulate you from the grunge, they
>often make it hard to obtain optimal performance. The set of fully portable
>modifications that yield pure improvement on any machine is very small
>compared to the set of modifications you can make when you have some
>knowledge of the machine hardware and how the compiler will react to the
>program (including the various optimizer switches you can invoke).

I've had some _very_ limited experience with the Convex here at Tulane, and
you're right: you have to write the code a certain way in order for the
vectorizer to be able to do much with it. Also, your point about I/O was
well taken.

However, as I've been following this discussion over the past couple of
weeks (and I've found it very interesting), I've sensed an overriding concern
for efficiency. While that's probably understandable in a newsgroup dealing
with a 2.8 Mhz home computer, I believe many times other factors tip the
scales against programming in assembly. For example, I'm sure that with
the revolving door mentality many people have about jobs, a lot of software
development houses put a premium on well modularized code with clean
interfaces that is easy to modify even by someone who didn't write it.
I don't deny this is possible in assembly, but I'd certainly say this is
easier to achieve in principle with an HLL.


>Writing portable code that compiles well on your machine is generally the
>best way to go; only when performance or coding time is more important than
>portability should assembly be used

Seems to me an HLL would be better when coding time is limited. I would say
it a little stronger: only when performance is more important than
_every_other_consideration_ should assembly be used.

Knowledge of assembly is certainly useful, which I'm sure is why it's
part of almost all computer science curriculums. But frankly, I think
there are a lot of considerations that make it an inappropriate choice for
most projects.

Ray
lang@rex.cs.tulane.edu