Quote:
Originally Posted by BetterRed
Lisp wasn't so bad, APL was a health hazard.
|
An old friend is an APL specialist. He was in the APL group at IBM at one point. He recounted an incident dealing with an APL vendor as a consultant where he said,
"I need to see your source code, so I can determine whether your implementation is fit for deploying at my client."
"Sorry, we don't provide source."
"That's okay. I'll disassemble your object code and look at the assembler listing."
"If you can do that, we'll give you a doctorate in computer science!"
"I already have one."
"Then we'll give you an office next to (APL creator) Ken Iverson!"
"I used to have one."
Game, set, and match.
He began in APL, moved to AStar, and now programs in an APL derivative called K (which does not require the special characters beloved of APL programming.) Most of his work is financial applications for which APL matrix operations are well suited, and he claims he can deliver completed applications in the time it will take developers working in other languages to determine how long it will take them to do what the RFC calls for.
I have no reason to disbelieve him.
As for Lisp, you get Greenspun’s Law: “Any sufficiently complicated C or Fortran program contains an ad-hoc, informally-specified, bug-ridden, slow implementation of half of Common Lisp.”
The underlying issue is whether the language you use includes automatic memory management, and if it doesn't, problems accumulate rapidly as the application increases in size and complexity.
______
Dennis