[Accessibility] FSGA Teleconference Agenda, Wednesday 28 June

Kris Van Hees aedil-access at alchar.org
Wed Jul 12 09:27:03 PDT 2006


Reading the performance analysis report, I really see quite some gaps in the
analysis that (IMHO) need filling out before the results can really be
validated.  For one, the code used to do the testing is not listed in its
complete form (only the interface descriptions).  Also, no in-depth analysis
is done of the results to support the conclusions on where the slowdown might
be located (in particular, there is no information about the server side
side implementation of the test code).  Finally, it seems to be implied that
whatever versions distributed with FC5 were used, though that is not mentioned
explicitly.  If so, I'd argue that without knowing the exact configure options
used to build the software involved, again, the analysis has a bit of a weak
foundation (from a purely scientific point of view).

That is not to say that the result may still be in favour of the CORBA
implementation used in Gnome.  But it would be great to see a more formal
analysis about the respective performance.  Especially if it may influence
future decisions on the AT architecture.

	Kris

On Wed, Jul 12, 2006 at 10:55:24AM -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.
> 
> Will




More information about the Accessibility mailing list