[dev] CVS session_destroy() and cookie-less sessions

[Chuck Hagenbuch]

Quoting mays@optonline.net:

> The patches below do have the side effect of calling session_start() twice
> (albeit with the intervening session_destroy()).  Perhaps it would be better
> to rearrange the flow a bit to clean it up ?  Or maybe session_start() needs
> to be called before session_destroy() ?

This seems to work for now; I've committed it. If someone has a cleaner 
solution we can use it.

-chuck

--
Charles Hagenbuch, <chuck@horde.org>
Some fallen angels have their good reasons.