Re: cvs: /php3 ChangeLog

On Wed, 24 Nov 1999, Sascha Schumann wrote: > On Tue, Nov 23, 1999 at 08:18:01AM -0500, Rasmus Lerdorf wrote: > > > or did i miss something? > > > > No, you didn't miss anything. Some bugs popped up and I have been slow in > > addressing them. The main thing left at this point is that the > > mysql_change_user() function needs to modify the hashed information for > > the persistent connection it modifies. > > This could be easily solved, if the MySQL module would not be > affected by a bad design decision which has been made for > many other modules as well. We can argue on whether this is a bad design or not, but it's basically a question of whether you keep things quick, or remain open for possible future improvements you can't really fortell. At the time, change_user() wasn't dreamt of, and keeping things as quick as possible was the number one priority. I don't think that there's anything holy about the way the MySQL module works when it comes to persistence. I wouldn't go about adding change_user() support by going over the list and rehashing afterwards, but rather, by modifying the entire architecture not to use user/password as a part of the key. It's not as if it's a big deal to change; Just as much, I wouldn't change it in databases that don't support the notion of change_user(). It's not a bad design issue (ok, so I'm arguing a bit :) I'll refamiliarize myself with my old code (and the new code) and modify it to this concept in the weekend. Zeev -- ----------------------------------------------------- Zeev Suraski <zeev@zend.com> http://www.zend.com/