>> "Geeklog needs the register_globals variable turned on in order to work.
>> Since PHP 4.2.0, the default for register_globals is "off". To fix it,
>> simply add the following line to your php.ini file
>Is this not *wrong*. It sounds to me like fake laziness. Nothing drives 
>me bonkers more than trying to track down a variable that is inherited 
>from who knows where. 
>Perhaps I'm missing something. I've recently worked on a lot of PHP 
>code written by others and it's a nightmare trying to track down where 
>a variable is defined and where it's value might be changed.  Perhaps 
>there is some tool I can use to trace this. I dunno. 
>Am I correct in my aversion to globals or I am I missing their true 
>value and perhaps some tools I could use when working on apps that have 
>more than 50 php files floating all over the place and no 

If it's not obvious where the value is coming from, whether they have
register_globals "on" or "off" isn't really going to help.

If the code is that messy, then they'll probably have something like:

  $i = $_REQUEST['i'];

in one file, and then start using $i out of the blue 37 files later, and 'i'
is *such* a great variable name, isn't it?

register_globals "off" only stops the over-riding of un-initialized
variables from the "outside" world (GET/POST/COOKIE).

If you've got crappy code with un-initialized variables in the first place,
you've got much bigger problems than register_globals -- You need to throw
out a big pile of code.

Unfortunately, some very popular but crappy code-bases out there continue to
be used by the masses, so we're all stuck with stupid register_globals
settings because of poor programming.

It's the right solution for a [bleeped]-up world, but that doesn't make one
not wish for a better world.

I guess I'm saying that there is only one correct tool to use for
applications with 50 php files floating all over the place and no

rm -rf application


Like Music?  http://l-i-e.com/artists.htm

PHP General Mailing List (http://www.php.net/)
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to