Re: D-Bus \approx Mbus

[Havoc Pennington <hp redhat com>]

On Mon, Mar 03, 2003 at 05:13:12PM +0000, Gustavo J. A. M.  Carneiro wrote: 
>   I don't agree.  This thing is originally meant for multimedia
> applications.  But it has messages, addressing, security,
> authentication, etc.  You shouldn't disregard it just because of the
> word 'Media'.
>   I realize you are in a better position to compare, since you
> implemented D-Bus while I implemented none.  But saying "not even
> remotely" is surely an exaggeration.

Any IPC mechanism has messages, addressing, security, and
authentication. ;-)

>   At least this thing has an RFC (http://www.ietf.org/rfc/rfc3259.txt),
> which makes it stand on its own.  Not that I'm advocating its use, but
> it has been thoroughly engineered, it is no longer ad-hoc, unlike
> D-Bus.

D-BUS is very engineered - it's third iteration, we've had DCOP used
widely and my earlier prototype FDMB, plus lengthy design discussions
with lots of people over a period of many months. Comprehensive specs,
automated test coverage, and complete docs are goals, and we've been
sticking to those goals pretty well so far. It is 100% API documented
and has excellent test suite coverage.

Havoc