Created attachment 3045 [details]
kgpg_CrashFullStart.patch

That is why the class KgpgAppletApp execution can be invoked with the "running" state but at missing "kgpg_applet" still.

Deriving from KUniqueApplication guarantees only one instance is running.
I'm not sure if it is not some local issue because you should have kgpg_applet != 0 when reaching this point.

Also why don't you come over to Gitea https://mirror.git.trinitydesktop.org/gitea/

And BTW I am not sure your patch is correct.

(In reply to deloptes from comment #2)
> Deriving from KUniqueApplication guarantees only one instance is running.
> I'm not sure if it is not some local issue because you should have
> kgpg_applet != 0 when reaching this point.

But that crashes stable!

> Also why don't you come over to Gitea
> https://mirror.git.trinitydesktop.org/gitea/

Firstly, my patches just laying there without applying also.
Secondary, I don't want to deepen in your GIT infrastructure of the patches deploying since I want only one stable branch.

> And BTW I am not sure your patch is correct.

That works for me, then it is correct.

(In reply to Roman Savochenko from comment #3)
> (In reply to deloptes from comment #2)
> > Deriving from KUniqueApplication guarantees only one instance is running.
> > I'm not sure if it is not some local issue because you should have
> > kgpg_applet != 0 when reaching this point.
> 
> But that crashes stable!

I can not confirm that, may be wait for few more weeks to get 14.1 as stable and we see if this happens.

> 
> > Also why don't you come over to Gitea
> > https://mirror.git.trinitydesktop.org/gitea/
> 
> Firstly, my patches just laying there without applying also.
> Secondary, I don't want to deepen in your GIT infrastructure of the patches
> deploying since I want only one stable branch.
> 

Freedom of choice is respected, but I am also not sure that someone is working on the bugs raised here, because all work moved to GIT.
I was just looking into some older stuff and saw, that you are still raising bugs here. But may be I am wrong and someone is taking care of them.
In any case it would work better if you raise bugs there.


> > And BTW I am not sure your patch is correct.
> 
> That works for me, then it is correct.

Well, keep in mind you are not alone on this planet. I think you should be looking to find out why it is not initialized, when you reach this point, because the code seems to initialize the object if !running.

BR

(In reply to deloptes from comment #4)
> I can not confirm that, may be wait for few more weeks to get 14.1 as stable
> and we see if this happens.

It crashes from the TDE beginning for me and that is why I did not started it by default. Why must I wait, when I use own patches constantly? :)
But sure, I am going to adapt all my patchset to 14.1 once its release.

> > > Also why don't you come over to Gitea
> > > https://mirror.git.trinitydesktop.org/gitea/
> > 
> > Firstly, my patches just laying there without applying also.
> > Secondary, I don't want to deepen in your GIT infrastructure of the patches
> > deploying since I want only one stable branch.
> 
> Freedom of choice is respected, but I am also not sure that someone is
> working on the bugs raised here, because all work moved to GIT.
> I was just looking into some older stuff and saw, that you are still raising
> bugs here. But may be I am wrong and someone is taking care of them.
> In any case it would work better if you raise bugs there.

That is not about freedom if you are poking the users into GIT, that is all users must be became developers of TDE and users who do not want, but have big experience, ignored in the fact. If Bugzilla is not actual, why don't close it or block new issues creation at least? Then I will support my own patchset without any notification. :)

> > > And BTW I am not sure your patch is correct.
> > 
> > That works for me, then it is correct.
> 
> Well, keep in mind you are not alone on this planet.

How "not alone" I know well from year 2012, when I start the TDE fixing and support it in the Linux distributive ALTLinux and now I support for much enough patchset already — http://oscada.org/wiki/Special:MyLanguage/Sub-projects/Automation_Linux_distributive . :)

> I think you should be
> looking to find out why it is not initialized, when you reach this point,
> because the code seems to initialize the object if !running.

This patch works FINE for me from the issue opening, so that is correct and all these were checked!

Created attachment 3063 [details]
kgpg.kcrash

Actual crash on 14.1.0!