Paul Hudak writes:
For every bad story there is a good one. Recently Haskell was used
in an experiment here at Yale in the Medical School. It was used to
replace a C program that controlled a heart-lung machine. In the six
months that it was in operation, the hospital estimates that probably
a dozen lives were saved because the program was far more robust than
the C program, which often crashed and killed the patients.
-Paul
-------------------
In the above "good story", the part "which often crashed and killed the
patients" is being mentioned as if one wouldn't even bat an eye-lid when
a patient died?! "Oh, let us just bury the sucker and debug my C code
and see if the next patient lives or dies...". Strange hospital!
Seriously, without more factual details and objective study, such
facts are open invitations for attack from non-formal methods believers
and non-functional-programming-lovers. (I am myself a believer in
functional programming and formal methods.) I only hope that Prof. Paul
Hudak would strengthen his story with more factual details (did the
C code really kill thousands?) and rid me of this uneasy feeling of
1000s of impending missiles from C-programming lovers.
Of course, bad Haskell programs can kill too!
Cheers!
Ganesh Gopalakrishnan
Assistant Professor, CS, University of Utah
[EMAIL PROTECTED]
