Re: How To Profile A Program

[Ralf Stephan]

Darren: 
> The steps you provide are very clear.  I am going to give it a try when I
> get over a couple of hurdles this week.  But, I'm still a bit unclear as to
> what one learns by profiling.

Generally (have a look at 'info gprof', and if noone has filled the
niche of a gprof book (search engine), then someone should), you get
the time the CPU spends in a function.  This may add up.  And I agree
with what was said about 'clearer/fine-grained design is better than
fast spaghetti'.  For 'stepping back' from the code to get more holistic,
pattern thinking is nice --- here a factory, there a bridge, facade etc.
Search for "Design Patterns", esp. the book by Gamma, Vlissides et al,
or, if you have time, check
http://www.laputan.org/mud/mud.html
http://www.laputan.org/foote/papers.html

I've also seen a raving review of "The Pragmatic Programmer" by Hunt/Thomas
on
http://slashdot.org/books/00/02/15/1912222.shtml


ralf
-- 
http://me.in-berlin.de/~rws/
If it's another exponential curve, in 2020 the world will have 20 launches/day.

                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.