From kde-core-devel  Thu Jul 13 18:44:04 2006
From: Waldo Bastian <bastian () kde ! org>
Date: Thu, 13 Jul 2006 18:44:04 +0000
To: kde-core-devel
Subject: Re: DBus/QtDBus Concerns
Message-Id: <200607131444.11338.bastian () kde ! org>
X-MARC-Message: https://marc.info/?l=kde-core-devel&m=115282708708989
MIME-Version: 1
Content-Type: multipart/mixed; boundary="--nextPart10878137.3WKSjgaRpD"

--nextPart10878137.3WKSjgaRpD
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Wednesday 12 July 2006 14:25, Reinhold Kainhofer wrote:
> Am Mittwoch, 12. Juli 2006 19:56 schrieb Olaf Jan Schmidt:
> > Hi Marius!
> >
> > On the FSG Accessibility list, people are questioning our planned move =
of
> > the AT-SPI infrastructure from CORBA to DBUS based on the analysis
> > already named by Gary:
> > http://eleceng.dit.ie/frank/rpc/CORBAGnomeDBUSPerformanceAnalysis.pdf
> >
> > Unfortunately the analysis does not really explain the exact bindings
> > used, so it is impossible to evaluate properly.
>
> All the executables, as well as the source code and the build files
> (either .sh scripts or Makefiles) are available from that page, too:
>    http://eleceng.dit.ie/frank/rpc/benchmarking.tar.gz
>
> Of course, they are using their own gnome corba/dcop/dbus bindings. We
> might try it out with kde's dcop and qt's dbus bindings, too.
>
> Also, he's just evaluating the speed of many function calls with
> practically no data transferred. I wonder how the results are when large
> data chunks are transferred.

Note that the explanation on the performance difference points to the=20
dispatching. That's something that is fixable in the binding I think, if=20
someone would really care. (or if someone really wants to make a lot of=20
calls)

Cheers,
Waldo
=2D-=20
Linux Client Architect - Channel Platform Solutions Group - Intel Corporati=
on

--nextPart10878137.3WKSjgaRpD
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2.2 (GNU/Linux)

iD8DBQBEtpR7N4pvrENfboIRAjrZAJ4skdCL+R8oNr8/4u5VCtykKvg4WACeOe2e
ifZ3qyzqrj7gMAEVQwNslUc=
=r5KY
-----END PGP SIGNATURE-----

--nextPart10878137.3WKSjgaRpD--