I think the issue is not compatibility with external libraries but
with previous versions of the engine.  IMHO, there is no reason for
a module in the HEAD branch to be compatible with older versions of
PHP.  In the case of MySQL and GD, the ifdefs are used to maintain
compatibility with the libraries, not PHP/Zend.

Just my $0.02

Sean

> -----Original Message-----
> From: Adam Dickmeiss [mailto:[EMAIL PROTECTED]]
> Sent: Sunday, February 10, 2002 12:30 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [PHP-DEV] Re: [PHP-CVS] cvs: php4 /ext/msession 
> msession.c
> 
> 
> On Sun, Feb 10, 2002 at 12:20:55PM +0100, Sebastian Bergmann wrote:
> > "Stig S. Bakken" wrote:
> > > Well, shouldn't Mark, who maintains the code in question, 
> be the one to
> > > decide whether a couple of extra ifdefs makes his code 
> unmaintainable?
> > 
> >   Well, needs msession a ''prime time seat'' in /php4/ext/, 
> or wouldn't
> >   we all be better off putting it into /pear/?
> 
> I certainly are one of those that don't like ifdefs. It's a bad sign
> and it's difficult to test - read - the code. Having said 
> that, msession
> have few ifdefs compared to many other modules. We could throw
> mysql and gd away  - they have more ifdefs! (pear is not a 
> garbage can.
> If it is treated as such, nobody will move).
> 
> What is interesting is that GD tops. I love GD but is one of the
> nastiest modules to ./configure.
> 
> Actually a few ifdefs that reflect the changes in Zend is not 
> that bad (what
> is bad is the change of the interface). At least the #ifdefs 
> force programmers' to
> watch out! Furthermore, code that do have ifdefs is easier to 
> move to pear.
> As for the documentation it's the same story. If an interface 
> changes over
> time AND that is reflected in the code (because of ifdefs) 
> they know that
> this interface have changed between PHP versions.
> 
> For code that is going to stay within ext I think we should 
> *allow* for
> ifdefs that make the code work with at least the previous stable
> version. The obvious good reason is testing. For example if somebody
> breaks apache apxs (which did happen), it's good that I can test code
> with the previous releasee to be able to continue coding on 
> the module.
> 
> -- Adam
> 
> > 
> > -- 
> >   Sebastian Bergmann
> >   http://sebastian-bergmann.de/                 
http://phpOpenTracker.de/
> 
>   Did I help you? Consider a gift: http://wishlist.sebastian-bergmann.de/
> 
> -- 
> PHP Development Mailing List <http://www.php.net/>
> To unsubscribe, visit: http://www.php.net/unsub.php

-- 
Adam Dickmeiss  mailto:[EMAIL PROTECTED]  http://www.indexdata.dk
Index Data      T: +45 33410100           Mob.: 212 212 66

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



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

Reply via email to