Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!super!super.org!rminnich
From: rminnich@super.org (Ronald G Minnich)
Newsgroups: comp.arch
Subject: Re: Compilers taking advantage of architectural enhancements
Message-ID: <36493@super.ORG>
Date: 2 Nov 90 19:28:11 GMT
References: <1990Oct9> <3300194@m.cs.uiuc.edu> <11922@ganymede.inmos.co.uk> <8424@scolex.sco.COM> <2694@l.cc.purdue.edu> <1990Oct31.203932.26325@cs.cmu.edu>
Sender: news@super.ORG
Reply-To: rminnich@super.org (Ronald G Minnich)
Organization: Supercomputing Research Center
Lines: 23

In article <1990Oct31.203932.26325@cs.cmu.edu>, spot@WOOZLE.GRAPHICS.CS.CMU.EDU (Scott Draves) writes:

|> name it.  and then show how it can be implemented on a modern processor.
|> and then show that those transistors couldn't have been better spent on
|> something else.  keep in mind the application mix that uProcs are
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
|> actually used for.
^^^^^^^^^^^^^^^^^^^^^

I understand your point, and this is an old argument, but it is beginning to 
sound parlously close to circular logic: 
A) I won't implement x because no one currently uses it
U) I can't use this uP, since it can't do x, better go use something else
   (two years later)
A) See what i mean? nobody uses x!

ron

-- 
"Socialism is the road from capitalism to communism, but we never promised to 
                 feed you on the way!"-- old Russian saying
"Socialism is the torturous road from capitalism to 
                  capitalism" -- new Russian saying (Wash. Post 9/16)