Re: var renaming, speed testing and no-cache

--- Monte Ohrt <monte@ispi.net> wrote: > Since the abstraction was aimed at speeding up instantiation time, I > didn't bother benchmarking internal execution times, those should > largely be the same, unless other changes (outside of abstraction) > have affected them. I see. I certainly agree that it is important to have quicker startup times, especially on very busy sites, so this is good news. @wom.bat: going to the Canadian west coast -- Vancouver, Whistler and Hornby Island. yee haw! -- I know this is the wrong list, but ... as a general PHP rant, one thing that bothers me is that on Windows, PHP will load file A.php whether you do a require_once "A.php" or require_once "a.php". BUT, PHP's internal caches treats each file as a seperate file so that if A.php contains a class definition, then using both forms of the above leads to PHP thinking that you are trying to redeclare your class! I always thought that case-sensitive file names were one of the stupider things that Unix did and I am sad that PHP has extended it to its file caches. It may have some merit on nix boxes but it is non-intuitive on windows boxes. It gets worse now that PHP has stream support so that if I try to load http://mysite/A.php, it REALLY thinks that is a different file, so again, it tries (and fails) to redeclare the class. I would rather PHP through a warning instead of terminating on class redeclarations. Or perhaps if it CRC'd the file if it hit a redeclaration error to see if it really is just the same file again. PHP has the strangest/worst set of casing rules of all modern languages, mainly because it is so inconsistent. Any idea where it is best to post these rants, er, concerns?? I'd post to internals, but they are a bunch of sharks there. __________________________________ Do you Yahoo!? SBC Yahoo! DSL - Now only $29.95 per month! http://sbc.yahoo.com