home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
ftp.cs.arizona.edu
/
ftp.cs.arizona.edu.tar
/
ftp.cs.arizona.edu
/
icon
/
newsgrp
/
group00b.txt
/
000011_icon-group-sender _Fri Jul 7 14:21:52 2000.msg
< prev
next >
Wrap
Internet Message Format
|
2001-01-03
|
3KB
Return-Path: <icon-group-sender>
Received: (from root@localhost)
by baskerville.CS.Arizona.EDU (8.9.1a/8.9.1) id OAA21021
for icon-group-addresses; Fri, 7 Jul 2000 14:21:44 -0700 (MST)
Message-Id: <200007072121.OAA21021@baskerville.CS.Arizona.EDU>
From: gep2@terabites.com
Date: Fri, 07 Jul 2000 16:28:44 -0500
Subject: Re: Error messages
To: icon-group@optima.CS.Arizona.EDU
Errors-To: icon-group-errors@optima.CS.Arizona.EDU
Status: RO
Content-Length: 2450
> I think this is where things come together: Different languages for different
parts of an application.
Maybe you should build a vehicle which is simultaneously a boat, an airplane, a
car and a submarine, too. That way you can meld those disparate parts into a
cohesive (cough!) whole, too. ;-)
> Icon would have its place here for analyzing the input - maybe as far as to
storing tokens in a tokenarray.
To suggest that Icon's appeal for writing an assembler is limited only to the
initial syntactical parsing of tokens is pretty ludicrous.
> Maybe an assembler was a bad example - but still: Wouldn't ProLog be ideal for
making decisions?
No, not if those decisions are relatively simple... that's like using atomic
weapons to kill mosquitos.
Icon's goal-directed evaluation can be a wonderful tool for analyzing various
possible instruction sequences and rating them based on size, performance, and
other criteria. Icon also makes it hugely simpler to create all manner of
statistics and summaries and other information (and present it nicely) regarding
the program being processed, too.
> And then there is another matter: The instructions must be provided to the
processor (Pentium) in the optimal sequence, making sure the pipelines are full
at all times (as far as possible).
That was more of an issue in the days when there was one overwhelmingly dominant
microprocessor out there. With AMD, Cyrix, and others out there (even several
quite different-internally designs from Intel) it is largely impractical to
optimize for a specific microprocessor's internal architecture (unless you're
doing an embedded app for a specific KNOWN CPU architecture, such as a TI DSP or
something). And if you were GOING to optimize for just one processor, I'd think
it would make more sense to do it for an AMD processor, since I believe that
with recent stupid moves by Intel, I think you're going to see a lot of market
share slip out of Intel's hands and into those of AMD.
> Even the GNU tools haven't fully kept up here, since this is a big job, and,
in my mind, C is not the ideal tool for this job.
C isn't ideal for most anything that isn't really low level.
Gordon Peterson
http://web2.airmail.net/gep2/
Support the Anti-SPAM Amendment! Join at http://www.cauce.org/
12/19/98: the day the Conservatives demonstrated their scorn for their
fraudulent sham of representative government. Voters, remember it!