[maemo-developers] osso/d-bus bug?

From: Charles Werbick roadmapformaemo at gmail.com
Date: Mon Jan 14 20:13:22 EET 2008
Hello All,

This is my first post to this list. I've scoured the internet and the
mailing list archives for info on this and have only found one
unconfirmed bug report regarding it. So- Apologies in advance, if it's
already been answered.

I'm a long time linux hacker who recently got the n810 for work
(rdesktop+ssh+vnc==never having to lug around a laptop again ;-)

I am currently porting roadmap (http://roadmap.sourceforge.net) to
maemo with great success. Hildonization and deb packaging are a
breeze. But I've run into issues with d-bus service implementation
using osso. Here's the thing-

I am able to cause an abort to dbus from userspace when starting the
app from maemo-launcher! On the device this kills then restarts the
entire ui. (Minus the desktop applets which do not reload.)

Note-When starting from the command line, everything works great.
dbus-monitor shows the service, and once the app is started, clicking
on the launch icon switches to the running app rather than start a new
instance(as it should).

I haven't finished the callbacks yet, which is the likely cause.
However,  it seems like osso and/or dbus should handle the situation
more gracefully than it does.

My questions-
1. Has anyone else observed this behavior?
2. Why does d-bus abort rather than just kill the app?
3. What are the minimum callbacks that I need to provide in order to
prevent this?

Any feedback would be appreciated.

Thanks in advance,
Charles Werbick

More information about the maemo-developers mailing list