Re: [Bug-xnee] xmodmap stalls xnee --record

[Henrik Sandklef]

Argghhh,

 there's already a work around in Xnee. Adding one more doesn't make me 
feel good.

 But first of all, thanks for the report and for your work on tracking 
the thing down :)

 We think one of the old synch bug is now fixed so chances are we 
proceed with this modmap thingie. But first a 3.03 (release process was 
halted due to the synch bug)

/hesa


Gelineau,Samuel [Montreal] wrote:
> I've updated to 3.02 and ran the command you suggested. I tried a couple of 
> scenarios and the last events reported before cnee stalls were invariably:
>
> 1,55,55,5,48234496,2773018735   (X_CreateGC)
> 1,98,98,5,12,2773018735         (X_QueryExtension)
> 1,20,20,6,341,2773018735        (X_GetProperty)
> 1,98,98,5,48234505,2773018735   (X_QueryExtension)
> 1,119,119,1,48234505,2773018735 (X_GetModifierMapping)
>
> I guess one way to patch the problem would be to detect this particular 
> sequence of events as being xmodmap's signature, and to print a message when it 
> is encountered. I think the last line might even be sufficient; I've found 
> other programs which generated X_GetModifierMapping messages, but none with 
> that particular value in the 5th field.
>
> Alternatively, one could arrange for this message to trigger a breakpoint, in 
> order to discover more about the problem and devise a cleaner solution.
>
> Good luck,
>
> - Samuel
>
> -----Message d'origine-----
> De : Henrik Sandklef [mailto:address@hidden 
> Envoyé : 11 avril 2009 18:45
> À : Gelineau,Samuel [Montreal]
> Cc : address@hidden
> Objet : Re: [Bug-xnee] xmodmap stalls xnee --record
>
> hi
>   and thanks for your report.
>
> You're using an old version of Xnee. The name xnee for the program hasn't been 
> there for a while.  Can you update to 3.02?
>
> ... and then do:
>
>   cnee --data-to-record 100 --record --device-event-range 2-6 
> --delivered-event-range 7-31 --request-range 1-120 --error-range 1-17
>
>
> We can now reproduce your reported error.... and to be honest I have no 
> idea why this happens.
>
> There's an old synch bug which really has to be solved before this 
> one.... hopefully soon.
>
>
> /hesa
>
>
>
> Gelineau,Samuel [Montreal] wrote:
> > Hi,
> >
> > You may or may not consider this a bug, but I've figured out that calling 
> > xmodmap while xnee is recording will cause xnee to stop outputing new events.
> >
> > How to reproduce:
> >
> > $ xnee --k-log --record --all-events > foo.log &
> > $ wc -l foo.log
> > 82 foo.log
> > $ wc -l foo.log
> > 85 foo.log
> > $ xmodmap -e 'clear Lock'
> > $ wc -l foo.log
> > 136 foo.log
> > $ wc -l foo.log
> > 136 foo.log
> > $ killall xnee
> > $ rm foo.log
> >
> > I personally don't see this as being very problematic, but I would have 
> > preferred xnee to halt or output a message instead of stalling. I also wanted 
> > to point out this behaviour in case other people saw their long-running xnee 
> > sessions hang mysteriously and wondered what could cause it.
> >
> >
> > - Samuel Gélineau
> >
> >
> >
> >
> > ------------------------------------------------------------------------
> >
> > _______________________________________________
> > Bug-xnee mailing list
> > address@hidden
> > http://lists.gnu.org/mailman/listinfo/bug-xnee