> The other advantage is to force people one way or the other. In the case
> of 50% of servers allowing short tags, and the other not... a script using
> short tags will only work on 50% of PHP installations (just as a script
> that relies on register_globals will only work on servers with it switched
> on).

Yes, but I don't particularly care what 50% of the servers out there do if
I am writing code to solve my problem on my servers.

> The only way around that problem is:
>  1. to force short tags on everywhere
>  2. to force people to use a tagging which is available everywhere

No, the way around that problem is to make it configurable.  As soon as
you force someone to do something, you are imposing a restriction which is
going to make it harder for him to solve his problem.

As you illustrated, there are some fringe cases where you may want to mix
PHP and XML, and in those cases the tool needs to be flexible enough to
still be useful.  If we forced short tags on, then you couldn't have
solved your problem using PHP, so option 1 is out.  Option 2 forces people
to conform to other peoples' server configurations and to at the same time
use much more verbose tagging for the sole reason of making their code
work on those other people's servers.  If those people have absolutely no
inclination to give out their code, why in the world do you want to force
them to jump through hoops?

We have done our job here. We have educated people and told them that for
maximum portability they should use <?php over <?. Forcing the issue
further is just going to piss people off without solving any real problem.

To continue down this path you have started, there are html authoring
tools that want all scripting inside <script language="php">...</script>
tags.  This tag-style is always on in PHP and is even more portable than
<?php ... ?>.  By your logic this is the tag-style we should be using
then?  Because if we don't, then there are people who will not be able to
edit your scripts in their environments.

-Rasmus


-- 
PHP Development Mailing List <http://www.php.net/>
To unsubscribe, visit: http://www.php.net/unsub.php

Reply via email to