Re: The joy of FORTRAN

On 3/1/25 3:29 PM, Charlie Gibbs wrote:

On 2025-03-01, Peter Flass <peter_flass@yahoo.com> wrote:
 

c186282 <c186282@nnada.net> wrote:
>

On 2/28/25 7:39 AM, Bob Eager wrote:
>

On Fri, 28 Feb 2025 02:34:52 +0000, rbowman wrote:
>

Sadly, macros made their way into C. The basic idea was fine but we had
one programmer who laid awake nights thinking up new perverted things to
do with macros. I forget if it was Kernighan or Ritchie who later said
if they had only known what people would do with them...

>
I probably do far too much with a standalone macro processor...

 If you're doing it for sport, though, there have been lots of
fun things done with macros.  My favourite is an IBM 360 macro
that would solve the Towers of Hanoi problem, and use the MNOTE
directive to show the moves.  It generated no executable code -
the expanded source code listing was the solution.
 I've written a couple of macro processors, and the acid test
for each was to write a recursive Towers of Hanoi macro for them.
 

   It's unfortunately NOT hard to make 'C' *totally
   unreadable*, impossible to follow. Sometimes you
   see the young hot-shots cram a dozen lines worth
   of stuff into a single line with LOTS of punctuation
   characters and bracket tricks. It's also possible to
   obfuscate using too many, esp nested, macros.
>
   Thing is, those ultra-compacted lines don't really
   run any faster - the same functions and ops are
   being run regardless.

 Again, if you're into it for sport, some amazing things
have been done, culminating in the International Obfuscated
C Code Contest:
 https://www.ioccc.org
 

   Being kinda old-school my pgms always look like
   something from the K&R book, and with comments
   after almost every line. THAT you can read a
   day from now, or a year from now ....

>
Never can have enough comments. I usually use lots, but recently I was
looking over some of my old code and had a “what?” moment. It was obvious
when I wrote it.

 I write C the same way.  A wise person on one of these newsfroups
once said that the comments aren't for someone else, they're for
you a year from now.  The more tricky the code, the more detailed
the comments.

   Good advice sadly FEW are willing to take.
   "Well, *I* understand it perfectly (today) !"
   Even 'clear' C is often NOT so clear - neither
   the syntax or logic. You've gotta EXPLAIN every
   step to yourself. Old code almost always has to
   be revised/debugged. Might be a month, a year,
   five years. How the hell DO you figure out what
   you, or anybody, worked 5 years along ? GREAT
   COMMENTING is the only way.
   Every function I wrote had a nice long explain
   at the top, then every line narrated in the
   body. It's the only thing that works.
   Some other langs are a little more self-documenting.
   Pascal is very readable. Python CAN be. COBOL was
   SUPPOSED to be. But even then the exact plan/meanings
   can be LOST after even a short time. The more
   clever the code, the worse it is.
   What's the point in writing complex code if, at
   the first need, you basically have to totally
   re-write the entire damned thing ???
   Oh well, 20 years, only the "AI"s will be
   doing 'programming" ..........