Re: a question of philosophy

2002-07-17 Thread
The FreeRADIUS authorization process retrieves the attribute information needed to perform the authentication process. IE, retrieving a password, setting the auth-type to use CHAP, PAP, EAP, etc. You can't authenticate the user until you know how you are supposed to authenticate them.

Re: a question of philosophy

2002-07-17 Thread Alan DeKok
=?utf-8?B?6ICA6YC4?= [EMAIL PROTECTED] wrote: so the authorization in the FreeRadius means gather the information for authentication, am I right? Yes, but it does more than that, too. It also gathers the authorization information (IP address, etc.) that the server *may* return to the NAS

a question of philosophy

2002-07-16 Thread Vic Abell
I'm new to the Radius protocol, just having finished implementing a module for access to a private authentication service. During development one thing struck me as odd: authorization checks are done before the entity being authorized is authenticated. It's been my experience that before an

Re: a question of philosophy

2002-07-16 Thread 3APA3A
Dear Vic Abell, Imagine you're coming to your president's room Secretary: do you have an appointment? Mr. Abell: Yes, my name is Vic Abel Secretary gets your name and looks into timetable and finds required record (that's authorization is). Than she checks time and name are

RE: a question of philosophy

2002-07-16 Thread Vic Abell
ZARAZA writes: Dear Vic Abell, Imagine you're coming to your president's room Secretary: do you have an appointment? Mr. Abell: Yes, my name is Vic Abel In this new an suspicious age, that wouldn't be the exchange. It would be: Secretary: Do you have an appointment? Mr. Abell:

Re[2]: a question of philosophy

2002-07-16 Thread 3APA3A
Dear Vic Abell, --Tuesday, July 16, 2002, 5:53:45 PM, you wrote to [EMAIL PROTECTED]: Secretary: do you have an appointment? Mr. Abell: Yes, my name is Vic Abel VA In this new an suspicious age, that wouldn't be the exchange. VA It would be: VA Secretary: Do you have an appointment? VA

Re: a question of philosophy

2002-07-16 Thread Alan DeKok
Vic Abell [EMAIL PROTECTED] wrote: During development one thing struck me as odd: authorization checks are done before the entity being authorized is authenticated. Yes, by design and intent. It's been my experience that before an entity is authorized it should be asked to prove itself

Re: a question of philosophy

2002-07-16 Thread Alan DeKok
Vic Abell [EMAIL PROTECTED] wrote: I don't think someone should be authorized before the claimed identity has been authenticated. Otherwise authorization might be given to someone falsely claiming an identity. Nonsense. The authorization isn't returned to the caller until after they've

RE: a question of philosophy

2002-07-16 Thread Vic Abell
Alan DeKok writes: Vic Abell [EMAIL PROTECTED] wrote: I don't think someone should be authorized before the claimed identity has been authenticated. Otherwise authorization might be given to someone falsely claiming an identity. Nonsense. The authorization isn't returned to the

Re: a question of philosophy

2002-07-16 Thread Alan DeKok
Vic Abell [EMAIL PROTECTED] wrote: Nonsense. The authorization isn't returned to the caller until after they've been authenticated. No, it's not nonsense. The secretary's telling me that Vic Abell has an appointment gives away potentially useful information. Please read again,

RE: a question of philosophy

2002-07-16 Thread Vic Abell
Alan DeKok writes: Uh, right. Why were you arguing about something you didn't understand? It would have been politer for you to ask HOW it works, rather than claiming it's wrong and insecure, and then back-pedalling when your confusion was corrected. Well I don't think I ever said