[Thunar-dev] Mounting fails due to no available org.freedesktop.Hal services

[Jannis Pohlmann]

On Sat, 13 Jan 2007 10:04:06 +0100, Benedikt Meurer wrote:

> Jannis Pohlmann wrote:
> > On Fri, 12 Jan 2007 22:40:10 +0100, Benedikt Meurer wrote:
> > 
> >> Jannis Pohlmann wrote:
> >>> But: Whenever I start thunar with a running hald (sorry, my last
> >>> mail contained a lot of hald-* processes, but not "hald" itself)
> >>> and try to mount one of the devices, I get this:
> >>>
> >>> jannis at nebelsee ~ $ thunar
> >>> Thunar: Failed to connect to the D-BUS session bus: Failed to
> >>> connect to socket /tmp/dbus-TTavTtFLD4: Verbindungsaufbau
> >>> abgelehnt
> >> D-Bus system daemon not running.
> >>
> >>> And after that, "hald" is not running anymore (but the helper apps
> >>> are, as listed my last mail). Weird, isn't it?
> >> Well, either a problem with D-Bus or a problem with HAL.
> > 
> > Probably, yes. It's just that I can't really verify what's wrong.
> > I've D-BUS 1.0.1, D-BUS GLib bindings 0.72 and HAL 0.5.8.1
> > installed on my system (Lunar Linux, so none of the original config
> > files - like /etc/dbus-1/system.d/hal.conf - were modified).
> > 
> > Now if I start D-BUS and HAL and try to mount something, I get this
> > "... is not provided by any .service files" error message. After
> > that, hald isn't running anymore. This is what happens as root. Now
> > if I set up D-BUS to give normal users access to some HAL-related
> > D-BUS services, like you suggested to Samuel in his bug report, I
> > get exactly the same error.
> > 
> > I know the package versions are pretty much bleeding edge as Lunar
> > isn't bound to fixed package release cycles, but nonetheless I'd
> > love to be able to get some more debugging information out of
> > Thunar/HAL/D-BUS to find a fix for this annoyance. After all, I
> > remember that it worked before exo-mount was introduced.
> 
> HAL is probably crashing in the Mount method. Run hald with
> --daemon=no --verbose=yes.

Believe it or not, but when I tried to run "hald --daemon=no
--verbose=yes", it immediately worked. And on the next day, when I
wanted to look into why it didn't work with the hald init script, that
also worked.

I don't really like problems randomly disappearing because this gives
me a feeling of not having control over them. Anyway, problem solved,
no need to invest more time into searching a bug that does not exist.

Thanks,
Jannis