Re: Authentication vs. Authorization
| From: | \[php\]Walter | Date: | Wed, 12 May 2004 17:58:37 +0000 |
| Subject: | Re: Authentication vs. Authorization | ||
| Groups: | php.pear.general | ||
| Request: | Send a blank email to pear-general+get-12436@lists.php.net to get a copy of this message | ||
Martin Jansen made pretty good arguments in the past in regards to
Authentication vs Authorization...
http://marc.theaimsgroup.com/?l=pear-general&m=100252782901898&w=2
SUB: Re: [PEAR] a user class
DATE: 2001-10-08 7:58:54
http://marc.theaimsgroup.com/?l=pear-general&m=100893621520427&w=2
SUB: Re: [PEAR] Pear Auth
DATE: 2001-12-21 12:03:09
Until I actually sat down and read ALL the threads in PEAR:GENERAL about
Auth, I too had the thought that Auth should have some Authorization method
to it.
But, after some thought, and reading the history here, I've come to agree
that Auth should stick to what it does best and something else can handle
Authorization.
I've even taken this a step farther and have concluded that addUser() and
remove User() should not even be in PEAR:Auth. But it might be a little late
for that now.
Now, with the new auth in CVS, I see that changePassword() has been added.
(Yes, I know, I asked for it some time ago. But I've "seen the light!")
Since Martin and a few others convinced me that the route to take is to
create an extension to Auth to handle authorization, have spent the last few
months, on and off, doing just that.
AuthUser is mostly complete, it handles the complete gambit of operation for
authentication and authorization. It builds off Auth to handle the
authentication side, and then it handles the authorization. There could be
containers to handle the various backends (I can't test against many, so I
will need some help there), just like Auth. All I have is DB right now.
(Yes, I know LiveUser exists)
One thing I do have in AuthUser, and I think it should be in Auth is...
// How many times will we allow a login attempt?
// ZERO (default) - no limit
//$objAuth->setLoginAttemptsMax(4);
If a user fails the given times, login will not happen again.
It needs to have something to stop a user from simply closing a browser and
trying again. But, it's my first stab at this. I figure the security gurus
here could solve this issue in about 10 seconds.
Anyway. I would appreciate some feedback on this.
Thanks
Walter
======================================================================
Torres Development Group Systems Design & Development Services
web.torres.ws C++ Perl PHP
V:708-288-5859 F:714-908-8403 [d]HTML CSS Javascript
======================================================================