🚀 go-pugleaf

In article <46oll9$3sm@reuters2.mitre.org>, "W. Wesley Groleau" <wgroleau@lear.mitre.org> writes:
> mfeldman@seas.gwu.edu (Michael Feldman) wrote:
> >(I)   It's easier to make a correct program fast than a fast program correct;
> >(II)  Fast enough is fast enough;
> >(III) If you've _measured_ it and it isn't fast enough, do what you have
> >      to do to get the performance up;
> >(IV)  90% of the work is done by 10% of the code, so don't bother
> >      optimizing the other 90% of the code.
> 
> Unfortunately, many people believe that
> (I)   It's easier to make a correct program fast than a fast program correct;
>       but the latter is what I have to do.

In that case, write a correct program and speed it up.  Both the program and
development time will be faster in the long run!  

> (II)  There's no such thing as fast enough;

In that case, you are in a permanent no-win situation (ongoing?).
Whatever you do will be inadequate, so why bother at all?

> (III) If your job description includes the words "real-time" do anything
>       you feel like to get the performance up--especially

Definitions:
Real-Time		When there are 60 seconds in a minute

REAL Real-Time		When there are 1000 milliseconds in a second

REALLY Real Real-Time  	When there are 1000 uS in a millisecond

> (IV)  turn off the optimizer and butcher^H^H^H^H^H^H^Hoptimize the code
>       until it screams.

Who screams?  You or the code?

Regards to all

Bob
~    ~~~  ~~~~	    ~~   ~~~  ~~ ~~   ~~  ~~~~~  ~~~	 __	~~  ~~  ~~  ~
                                   		       _<Zx\	  _ -/
     Bob Lang			              		   ||  _-  -/  	_ 
     lr-lang@csm.uwe.ac.uk		                  // _- -_ -/_- /    
     University of the West Of England	               	 //_- _ - _/- =/     
					~~~~~  ~~~~~~~~ (_-_-_- =_-===_/ ~~~~~
~~~ ~~~~~~~~~~ ~~~~~~  ~~~~~~~~~   ~~~~~~~ ~~~~~~~ ~~~~~~~  ~~~~~  ~~~~~~~ ~~
  ~~~~~~~~  ~~~~~~~~~~~~  ~~~~~~~~~~~~~~~    ~~~~~~~~~~~~~~~~      ~~~~~~~~ ~~