home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.edu:1572 comp.arch:9280 comp.lang.misc:2968
- Path: sparky!uunet!usc!rpi!bu.edu!wang!news
- From: amos@huji.ac.il (amos shapir)
- Newsgroups: comp.edu,comp.arch,comp.lang.misc
- Subject: Re: Scientists as Programmers (was Re: Small Language Wanted)
- Message-ID: <amos.715964119@shum>
- Date: 8 Sep 92 14:55:19 GMT
- References: <1992Sep3.112944.20996@dbsun.uucp> <Bu08uF.HBC@mentor.cc.purdue.edu> <amos.715786635@shum> <Bu64t4.B15@mentor.cc.purdue.edu>
- Sender: news@wang.com
- Reply-To: amos@cs.huji.ac.il
- Followup-To: comp.edu
- Organization: Mail to News Gateway at Wang Labs
- Lines: 25
-
- hrubin@pop.stat.purdue.edu (Herman Rubin) writes:
-
- >One of my students came up with the following, which is a much more
- >common situation than most believe. It is desired to read an element
- >from a buffer location, with an interrupt if the buffer empties. Now
- >this is very much like reaching a cache page boundary, but a user
- >program is needed to correct this exception. Since the hardware
- >for handling this, and some of the software, is essentially already
- >in place, could it not be extended? With the present setup, only with
- >considerable difficulty; at design time, easily.
-
- Not as easily as it looks to a programmer (I admit I also thought this
- way before working with the designers). In this age of multiple
- parallel units and pipelining, designers can save a *lot* of run time
- and chip space if the software people - mainly compiler writers - can
- assure them users will not use certain features at certain
- circumstances. Now, which features get left out is a matter of
- tradeoffs and some politics, in which small users (e.g. statisticians)
- lose out; if you want to keep a feature in, tell the designers it's
- needed for COBOL... :-(
- --
- Amos Shapir Net: amos@cs.huji.ac.il
- Paper: The Hebrew Univ. of Jerusalem, Dept. of Comp. Science.
- Givat-Ram, Jerusalem 91904, Israel
- Tel: +972 2 585706 GEO: 35 11 46 E / 31 46 21 N
-