[Beowulf] Teaching Scientific Computation (looking for the perfect text)

Robert G. Brown rgb at phy.duke.edu
Tue Nov 20 15:24:55 PST 2007

On Tue, 20 Nov 2007, Peter St. John wrote:

> Eventually the progression from an assembler programmer writing "Load
> Register 1 with 2" to an engineer writing "X = 2" will evolve to a
> business person asking "what are the projections for third quarter
> 08??!!".  It keeps getting easier for people expressing themselves

It already has.  But he doesn't ask a compiler, he asks a program.  A
compiler is the thing that CREATED that program, and there is an
interesting information theoretic argument that shows how it couldn't
create the program without the information required to create it already
being there, and that information does not come from the computer, it
comes from the programmer.

> more and more in the language of their subject, and less and less in
> the language of the computer, because the computer can learn to
> translate and interpret us. But I'm a wild-eyed optimist about AI.
> Also, I think it will be as easy to tell your boss "the compiler
> misunderstood me" as it is now to say "there's a bug" (if you aren't
> the programmer) or "it needed a pass thru QA" (if you are).

Arrrrgh.  Evil!  Back! Back!

Intelligent compilers will come AFTER we make intelligent computers, and
what will we use to give computers intelligence in the first place?

> Peter
> Incidentally, I'm a bit saddened that the GPL version of MACSYMA is
> "Maxima". Small thing.

Trademark, I'm sure.  The former probably belong(ed) to Digital or


Robert G. Brown
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone(cell): 1-919-280-8443
Web: http://www.phy.duke.edu/~rgb
Lulu Bookstore: http://stores.lulu.com/store.php?fAcctID=877977

More information about the Beowulf mailing list