[ILUG] introduction to programming

[Brian Foster]

  | Date: Wed, 25 May 2005 23:10:25 +0100
  | From: "John P. Looney" <valen at tuatha.org>
  | 
  | Pascal's only saving grace was that it compiled into pcode [ ... ]

 the UCSD(? Univ. Calif. at San Deigo) Pascal Sucks did,
 but others compiled to assembler/machine code.  the one
 I used generated executables directly.  that one was a
 fairly pure Pascal Sucks, lacking many of the common
 extensions that tried to turn the horrid pure language
 into something you might be able to use if the only
 alternative was a firing squad.   I _think_ that UCSD's
 Pascal Sucks had quite lot of extensions from Dr Wirth's
 original (pure) definition?

 amusingly, today I had to battle with a large Python
 script (on Linux, to keep this on topic) which generates
 a mixture of C and IDL code to test some parts of a
 distributed component system.  many parts of the script
 seem rather neat, but others are not.  I spent a good
 hour or three trying to track down why, went the script
 contained `foo' the generated files contained `bar'.
 finally found hidden away the equivalent of s/foo/bar/,
 in what design logic suggests is a backend routine that
 should do nothing of the sort.  *bletch*  ;-(   the
 goofy code seemed to be caused by some bad data design,
 which as said before in this thread, is at least as
 important as algorithm design — that design would have
 been just as bad in perl(1), C, &tc.  (judging by the
 comments, there was a requirements change that did not
 fit into the original design.  certainly the original
 authour is noted for usually writing quite good code.)

cheers!
	-blf-
-- 
Experienced (20+ yrs) kernel/software Eng: | Brian Foster   Montpellier,
 • Unix, embedded, &tc;  • Linux;  • doc;  | blf at utvinternet.ie   FRANCE
 • IDL, automated testing, process, &tc.   |  Stop E$$o (ExxonMobile)!
Résumé (CV) http://www.blf.utvinternet.ie  |     http://www.stopesso.com