[865] in Kerberos_V5_Development
Re: Proposed Kerberos V5 Password Changing Algorithm
daemon@ATHENA.MIT.EDU (Jonathan I. Kamens)
Fri Feb 24 12:04:47 1995
From: "Jonathan I. Kamens" <jik@cam.ov.com>
Date: Fri, 24 Feb 1995 12:07:25 -0500
To: krbdev@MIT.EDU
In-Reply-To: [863]
Date: Fri, 24 Feb 1995 00:11:00 EST
From: Marc Horowitz <marc@MIT.EDU>
In any case, Ted, it seems that there is "rough consensus" that the pw
change protocol should be text-based. The only person to disagree is
you....
I'm on Ted's side.
1) I don't think a text-based protocol is any easier to debug. If
you're monitoring the protocol for debugging purposes, you're going to
have to translate the armored ASCII into real data anyway, so if
you're writing a translator, your translator might as well handle the
binary data directly.
2) Writing code to convert the binary data into ASCII complicates the
task of writing a password server. I'll repeat Ted's point: This is
supposed to be a SIMPLE protocol, and it's supposed to be SIMPLE to
implement.
jik
