Newsgroups: comp.lang.apl
Path: watmath!watserv2.uwaterloo.ca!torn!cunews!csi.uottawa.ca!news
From: cbbrowne@csi.uottawa.ca (Christopher Browne)
Subject: Re: Promoting APL2
Message-ID: <1992Aug16.030509.19895@csi.uottawa.ca>
Sender: news@csi.uottawa.ca
Nntp-Posting-Host: prgv
Organization: Dept. of Computer Science, University of Ottawa
References: <168417F7D.JOHNSON2@vm.cc.purdue.edu> <1992Aug13.171145.24691@csi.jpl.nasa.gov> <1176@kepler1.rentec.com>
Date: Sun, 16 Aug 92 03:05:09 GMT

In article <1176@kepler1.rentec.com> andrew@rentec.com (Andrew Mullhaupt) writes:
>In article <1992Aug13.171145.24691@csi.jpl.nasa.gov> sam@csi.jpl.nasa.gov (Sam Sirlin) writes:
>>
>>In article <168417F7D.JOHNSON2@vm.cc.purdue.edu>, JOHNSON@vm.cc.purdue.edu writes:
>>I don't know what would end up being most popular, but for me there
>>are a few neccessities:
>>
>>1. linear algebra - as much of eispack/linpack/lapack as you can get.
>>Perhaps you could link them in via some sort of quad na? I've
>>translated the qr and qz eigensystem paths into APL if you want them.
>>I have them in readable ASCII form.
>
>NO! NO! A thousand times NO!. These guys have a 3090, and they should use
>ESSL, which is provided by IBM. Although ESSL has some trivial flaws, it
>is probably the only commercial package which is as good or better than
>LINPACK/EISPACK/LAPACK. ESSL has the world's _best_ FFT's bar none. (what
>else would you expect from Winograd et. al.?) Last time I used it, IBM
>had gone to the trouble to write the required interface for all of ESSL
>routines to quad-NA. Use the IBM provided ESSL.

I think you exaggerate a few times (maybe 995 or 996 :-)).

They may be using a 3090, but can you ASSUME that they have ESSL?  I
don't know all of IBM's marketing/pricing policies, but it's possible
that ESSL would be an expensive addition to the system.  (And in-house
development of an alternative would ALSO be expensive.  There may be a
price tag of $20,000 for ESSL, and a hidden price tag of $75,000 for
the in-house work.  Under different circumstances, one or the other
may be more "acceptable.")

IF they have it, it sounds like it's a pretty good idea to make use of
it.  But, it's not a universal panacea to all of the worlds problems.

For instance, it might be nice to develop an environment that's not
going to restrict you to 3090s.  Using {LIN,EIS,LA}PACK would allow
porting to other systems just a little easier.  This may not be the
FIRST thing they want, but using ESSL will prevent any sort of
portability.  Life is similar for the Runge-Kutta code.

-- 
Christopher Browne
cbbrowne@csi.uottawa.ca
University of Ottawa
Master of System Science Program
