Re: Re: cvs: php4 /ext/session session.c
| From: | mlwmohawk | Date: | Thu, 03 Jan 2002 13:30:14 +0000 |
| Subject: | Re: Re: cvs: php4 /ext/session session.c | ||
| References: | 1 2 3 | Groups: | php.cvs |
| Request: | Send a blank email to php-cvs+get-8919@lists.php.net to get a copy of this message | ||
On Wednesday 02 January 2002 04:37 pm, Yasuo Ohgaki wrote:
> Yasuo Ohgaki wrote:
>
> BTW, if I fix the issue with 3), I have to change msession a line
> so that it does not register it as session save handler. (handler
> is registered by #ifdef)
>
> I have to move pgsql session save handler to php4 source tree,
> since save handlers must be compiled with session module with
> option 3).
>
> Comments are welcome :)
A lot of mail comes through this list. Sorry if I cause a rehash. What I
don't understand is why there is the need to remove
php_session_register_module().
IMHO forcing the session module to know about other session handlers at
compile times is the wrong thing to do. I don't see why someone like RedHat
could not make a "supported" RPM release of PHP in which other developers
could write stand-alone shared library modules as save handlers.
Save handlers do not extened the PHP language and have a rigid interface.
Forcing a recompile and link just because you wish to add a few items to a
structure seems an absurd, almost Microsoftian, thing to do.
If we want to fix something that isn't working, we should fix it right. Let
Micrsoft hard code stupidity, we are better than that.