Re: From idea to project
Date: Fri, 6 Aug 93 09:01:39 -0700
Message-id: <>
Subject: Re: From idea to project
Sender: qed-owner
"No matter what we suggest, nothing is going to happen."
The fundamental reason for this is that our project, while appealing
for philosophical and theoretical scientific reasons, does not offer
short-term benefit to anybody.  I have just returned from a three-day
conference (a 70th birthday fest for Bob Finn) on minimal surfaces,
Navier-Stokes equations, and other topics in geometry and analysis.
Someone was quoted as saying "pde (partial diff eqns) is nothing but
inequality proving."  This refers to the ubiquity of "estimates" in 
which one bounds a hypothetical solution in some norm in order to 
prove existence.   But this still has to be done by hand, although 
one lecture was about checking some old calculations with Mathematica,
and several used computer graphics (Brakke's Evolver package) or 
numerics.   These people do not believe the computer will ever be 
able to assist them in their theorem-proving, and it is difficult to 
believe it for me too.   

I have a short-term suggestion of a semi-political nature.  Let's 
see if we can mobilize enough energy to get Math Reviews on-line on 
the Internet!  Here's the background:  MathFile, the MR on-line system,
is available on CD-ROM.  If your library has six thousand dollars you can
buy three CD-ROMS containing it.  This useful system lets you computer-search
Math Reviews 1980-now (although not all at once as you can search only
one disk at a time).   At the same time you can access MR on the Internet
but you can get only MR numbers, not reviews themselves and not even
bibliographical citations.  These CD-ROM systems in libraries whiiich 
have them are underutilized anyway (I'm not sure why, but people I've
spoken with say they would use it much more if it were on-line).  The 
AMS told me that they use the proceeds from the CD-ROM sales to support
the production of Math Reviews, and they are afraid that putting MR on-line
would kill the physical MR, which supports some eighty jobs in Providence.
Indeed it would:  MR should go all-electronic, be available ONLY or primarily
on the Internet, and be constantly updated, and paid for "somehow" by 
the AMS.  But since reviewers write for free, if it's in electronic form
it shouldn't cost much to add new reviews.