> So while we can debate computing considerations of today, tomorrow 
> those will be less important. That was the point I was making. Why 
> not focus on things that make significant difference and let the 
> insignificant fade into history.

I tendentiously focus on all things able to make, all together, even more 
significant difference.

Some micro-optimization, used as common code style, can make the entire 
application or the specific performance critical task, possible, even with an 
embed language as PHP is.

++$i is not different, from my point of view, from a code where each sequential 
push is performed via array_push($arr, $value) rather than $arr[] = $value;
Same is for all those loop such
for($i = 0; $i < count($staticStack); $i++);

for me alien, since I've always done

for($i = 0, $length = count($staticStack); $i < $length; ++$i);

or, even better, a core performed loop when I need values
foreach($staticStack as $value);

these are just examples, code style, whatever you want, and I'll never change 
my style unless there is a valid reason and some bench able to demonstrate I am 
wrong. I guess it's just a matter of point of views, but I cannot suggest 
slower practice cause Moore said tomorrow that CPU will strike the millisecond, 
'cause on micro benchmarks, we can go faster, and that's it.

Windows Live: Keep your friends up to date with what you do online.

Reply via email to