Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!lll-winken!uunet!ncc!alberta!cdshaw
From: cdshaw@alberta.UUCP (Chris Shaw)
Newsgroups: comp.arch
Subject: Re: Complex Instructions
Message-ID: <2249@pembina.UUCP>
Date: 3 May 89 23:16:36 GMT
References: <57252@yale-celray.yale.UUCP> <4101@tolerant.UUCP> <134@dg.dg.com> <504@daitc.daitc.mil> <1277@l.cc.purdue.edu>
Reply-To: cdshaw@pembina.UUCP (Chris Shaw)
Organization: U. of Alberta, Edmonton, Alberta, Canada
Lines: 50

In article <1277@l.cc.purdue.edu> cik@l.cc.purdue.edu (Herman Rubin) writes:
>In article <504@daitc.daitc.mil>, jkrueger@daitc.daitc.mil (Jonathan Krueger) writes:
>< >For example, we have a system with really wild instructions that gets
>< >generated even though the user don't know what they are and the
>< >compiler does only peephole optimization. As is turns out, having
>< >atomic operations for free is really nice.
>< 
>< Sounds like you've managed to let your users express what they want in
>< terms familiar to them but also understandable to the language
>< translator, following which optimization to particular hardware can be
>< done in a reasonable way.  By any chance, would this be done by way of
>< a function library?
>
>And what if your users find that the language is inadequate?  I know of
>no adequate one.  Do not cripple the programmer.

What if your stock Ford Escort doesn't do 180MPH on the straights? Don't cripple
the driver! What if your 2 micron CMOS gate array doesn't give 0.3 nanosecond
inverter propagation delay with fanout of 8? Don't cripple the logic designer! 
Get it?

The point I'm illustrating is that if you want to do something that 99% of the
rest of the world is unwilling to pay for, it should come as no surprise that
that the stock solutions fall short. Griping and complaining that the stock
solutions are inadequate is simply vacuous. If you invest the time and money,
you can get a Ford Escort to go 180. But it won't be stock. If you invest the
time and money, you can get 2.0 micron CMOS to give 0.3 ns propagation delay.
But it won't be a gate array, it'll be full custom.

Frankly, I'm getting very tired of reading Herman Rubin's complaints about C,
because he never seems to have a coherent solution. All I see is calls
for someone else to solve his particular problem, coupled with the unspoken
attitude that his particular complaint should have been taken care of long ago.
"The meter's running, buddy! Get to work!"

So what?

Maybe people do seriously listen, but come to the conclusion that it isn't
worth the time and effort to give Herman Rubin what he wants. I don't know, 
but it seems to me that Herman Rubin is wasting his breath. Time for action, I
think. Time for Herman Rubin to write "SL", for "Statistical Language", and
to report on its results. Calling other people stupid is valuable only as 
long as they are willing to listen. For me, time was up long ago.

>Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907

-- 
Chris Shaw    cdshaw@alberta.UUCP (or via watmath or ubc-vision)
University of Alberta
CatchPhrase: Bogus as HELL !