xfce4-session: xfwm4 removed from session, not started on login

Chris Bainbridge chris.bainbridge at gmail.com
Wed Jul 25 22:18:11 CEST 2012

On 25 July 2012 17:17, Olivier Fourdan <fourdan at gmail.com> wrote:
> Hi
> On Wed, Jul 25, 2012 at 1:13 PM, Chris Bainbridge
> <chris.bainbridge at gmail.com> wrote:
>> Hello,
>> I decided to try and track down the infamous bug that causes xfwm4 to
>> not be running on login (for details see links at end of email)
> Thanks for taking the time to investigate that, I could never reliably
> reproduce the issue and there was very little tangible data to get to the
> root cause.

I figured it was difficult to reproduce. I would experience the bug
about once a week in normal usage, I suspect it may have been more
common for me as I usually have gedit open and copy/paste work notes
into it, so my desktop would often be in a state where gedit would
have an unsaved document when I logged out, which would lead to the
failure and 60s timeout. Once the reason was known, the bug could be
reproduced almost every time with gedit running and an unsaved
document. I have also reproduced it by patching mousepad with the
patch that Dimitar Zhekov posted to bug #5379. I have tested both
normal and a cancelled logout with two interacts and both appear to
work with the user being presented both interactions.

> Yes, but such an issue should not happen, I don't see the point of adding a
> mechanism to prevent something that should not happen anyway.
> If xfwm4 crashes, it's automatically restarted, if another window manager
> takes control, it exits cleanly and the other window manager would be
> restarted by session management.
> If not, that's a bug that needs fixing.

Okay, I understand this line of reasoning.

More information about the Xfce4-dev mailing list