Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!wuarchive!udel!haven.umd.edu!uvaarpa!murdoch!astsun.astro.Virginia.EDU!gsh7w
From: gsh7w@astsun.astro.Virginia.EDU (Greg Hennessy)
Newsgroups: comp.arch
Subject: Re: Compilers and efficiency
Message-ID: <1991May10.150035.15438@murdoch.acc.Virginia.EDU>
Date: 10 May 91 15:00:35 GMT
References: <653@ctycal.UUCP> <28297C23.6984@tct.com> <12164@mentor.cc.purdue.edu>
Sender: usenet@murdoch.acc.Virginia.EDU
Organization: University of Virginia
Lines: 32

Chip Salzenberg writes:
#> That kind of problem is not "generally important" in that it does not
#> come up in the "general computing base".

Herman Rubin writes:
#That something is not important to everyone is no reason no to make it
#available to those who can see a use for it.

Chip didn't say something had to be important for everyone before
making it available, Chip said it had to be important to the general
base. Those aren't the same thing.

Herman, you seem to have quite specialized computer needs, and more
knowledge about computers than the average person, but the types of
things you do are not useful to J. Random User, so people who write
compilers for general computer users don't find it cost effective to
add your optimizations into their code.

As a radio-astronomer, I have specialized needs in computing also. I
don't complain that there is no commodity language to reduce my data,
instead I (and other radio-astronomers) write all our code ourselves.
Our market is not large enough to expect off the shelf solutions to
our problems.

Why don't you apply for a grant, and hire someone to write a language
for you?

--
-Greg Hennessy, University of Virginia
 USPS Mail:     Astronomy Department, Charlottesville, VA 22903-2475 USA
 Internet:      gsh7w@virginia.edu  
 UUCP:		...!uunet!virginia!gsh7w