[Accessibility] DBUS vs. CORBA (was Re: FSGA Teleconference Agenda,
Wednesday 28 June)
William.Walker at Sun.COM
Wed Jul 12 09:21:24 PDT 2006
It looks like the benchmarking code can be found here, though Frank
would be the best person to answer this question:
I want to make sure people know that I'm not pushing things one way or
another here. My only goal here is to understand if good decisions are
being made based on appropriate information.
On Wed, 2006-07-12 at 12:01 -0400, Jonathan Blandford wrote:
> [ adding John to the CC list ]
> On Wed, 2006-07-12 at 10:55 -0400, Willie Walker wrote:
> > > With respect to to the standard, we can always migrate from today's
> > > CORBA implementation to tomorrow's DBUS implementation
> > I've seen Frank Duignan's performance analysis work that provides
> > numbers to the effect that DBUS is approximately 18 times slower than
> > CORBA:
> > http://eleceng.dit.ie/frank/rpc/CORBAGnomeDBUSPerformanceAnalysis.pdf
> > I'm curious if there's been any response to Frank's work? I'd be
> > especially interested in what the DBUS developers have to say. Is the
> > analysis accurate/fair? Does the testing mechanism correlate to how we
> > use CORBA for the AT-SPI? Is there low hanging fruit in the DBUS
> > performance tree? Etc.
> Hi Will,
> That's a pretty interesting paper. Is the benchmark code available
> anywhere? Two immediate thoughts there are:
> 1. DBus on FC5 shipped with all the runtime message integrity checks on
> by default. The intention is to turn that off for DBus 1.0, but it adds
> an overhead right now.
> 2. That test is using the glib DBus bindings instead of the raw DBus
> interface. The glib bindings are known to be quite inefficient (lots of
> message copies), and are being rewritten.
> Of course, it could turn out that dbus is 17 times slower. It would be
> cool to have that benchmark to play with to make sure.
More information about the Accessibility