dbus first impressions

[Havoc Pennington]

On Tue, 2004-09-07 at 17:56 -0700, Don Park wrote:
> good point. what I should have said was, does the spec really mean to use
> SASL as an authentication mechanism, or is it specifying its own SASL-like
> mechanism. Sounds like the answer is the former.

Well, SASL isn't exactly a mechanism, it's sort of a meta-protocol; it's
kind of bizarre really. So what "uses SASL" means is that the D-BUS
custom auth protocol has the functionality described by the SASL spec.

> > On some old Unices, to send user/group credentials you have to send a
> > byte. The nul byte is this byte. I don't remember why it's nul rather
> > than 'a' or something, maybe there's no reason.
> 
> a dbus client app isn't talking to old unix daemons - you're defining both
> ends of the conversation, so why not leave the null byte out if the rest
> of the protocol is nice and keyboard-friendly?

This is an issue of the kernel, not the protocol or daemon. The way "get
credentials on other end of the socket" works on some UNIX involves
sending a byte.

Havoc