home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.arch:10876 comp.lang.misc:3765
- Newsgroups: comp.arch,comp.lang.misc
- Path: sparky!uunet!charon.amdahl.com!pacbell.com!decwrl!purdue!mentor.cc.purdue.edu!pop.stat.purdue.edu!hrubin
- From: hrubin@pop.stat.purdue.edu (Herman Rubin)
- Subject: Re: Hardware Support for Numeric Algorithms
- Message-ID: <BxxGKp.81A@mentor.cc.purdue.edu>
- Sender: news@mentor.cc.purdue.edu (USENET News)
- Organization: Purdue University Statistics Department
- References: <721627374@sheol.UUCP> <Bxpoy9.GD6@mentor.cc.purdue.edu> <722061187@sheol.UUCP>
- Date: Wed, 18 Nov 1992 19:53:12 GMT
- Lines: 71
-
- In article <722061187@sheol.UUCP> throopw@sheol.UUCP (Wayne Throop) writes:
- >: From: hrubin@mentor.cc.purdue.edu (Herman Rubin)
- >: Message-ID: <Bxpoy9.GD6@mentor.cc.purdue.edu>
- >::: Nobody except a computer programmer can read it at all!
- >:: The similar complaint that "those statisticians and their screwey
- >:: sums-over and chi-squares and their standard deviations! [...]"
- >:: [...] is false.
- >: Students in statistics courses have little problem with the notation
- >: you have mentioned
-
- >So, Herman agrees with me about statistics notations. I wonder what
- >reasons he might have to continue to DISagree with me about programming
- >notations, since students in computer courses have little problem with
- >the notations taught therein.
-
- The students in these statistics courses are very definitely NOT statistics
- students. Most statistics students are graduate students, and these methods
- courses are not required, nor even recommended.
-
- >: A biologist does not need so much to be fluent in statistical jargon
- >: to work in population genetics as to learn basic probability, and some
- >: basic statistics,
-
- >But if that biologist is going to talk to statisticians about novel
- >applications and novel methods, it *will* be necessary to learn more
- >than these rudiments. Similarly, if a statistician attempting to
- >help this biologist needs to develop new computer applications beyond
- >canned applications, that statistician will have to learn more than
- >is common about computer jargon.
-
- Novel applications are no problem. Novel methods are not going to be
- developed by biologists as biologists.
-
- The statistician will need numerical analysis jargon. That computer
- jargon diverges from this is essentially a move on the part of the
- computer language designers to distance themselves from mathematics.
- They have been too succesful in this.
-
- >::: And what is one supposed to do if the hardware operations wanted are
- >::: not even expressible in the language?
- >:: Use another language (or implementation).
- >: Bob Silverman and those working with him use huge gobs of computer time.
- >: They cannot get the language or implementation needed.
-
- >Nonsense. Assembly language *at* *least* has direct access to all of
- >the hardware operations. If no assembler exists for the machine, I'm
- >sure that one could be built quickly and cheaply. And with infix
- >operator syntax if desired.
-
- Dik Winter uses computers a lot more, and is quite accustomed to using
- assembler language. But he has posted that he is now faced with using a
- machine for which he does not have access to an assembler.
-
- My complaint about assemblers is that they have a syntax designed to be
- easy for the assembler, but hard for the human.
-
- >If Herman means he can't get a custom-designed language system for
- >peanuts, somehow I'm not surprised, and I doubt that the reason is
- >some conspiracy of language implementors to deny him the tools he wants.
-
- I do NOT want a custom-designed language system. What I want is a language
- which is flexible enough, and allows augmentations, so that one does not
- have to fight it. There is now no computer language which is not overly
- restricted. For numerical programming, C, if it had multidimensional
- arrays and exponentiation, would be superior in almost all respects to
- Fortran, and would become more widely used for that purpose.
- --
- Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
- Phone: (317)494-6054
- hrubin@snap.stat.purdue.edu (Internet, bitnet)
- {purdue,pur-ee}!snap.stat!hrubin(UUCP)
-