libsigcx-main

[Libsigcx-main] bug in GtkDispatcher::timer_callback()

[Klaus T. <tri...@ao...>]

hi,

while testing timeouting handlers in GtkDispatcher i found a bug (version  
0.6.4):

in file gtk_dispatch.cc, line 50, the event (with its callback) is erased  
 from the map, but the callback is called a few lines later - the handler  
gets never called because the slot object is not valid anymore.

--
http://triendl.info
klaus triendl

Re: [Libsigcx-main] bug in GtkDispatcher::timer_callback()

[Andreas R. <a.r...@gm...>]

"Klaus Triendl" <tri...@ao...> writes:

> hi,
>
> while testing timeouting handlers in GtkDispatcher i found a bug
> (version  0.6.4):
>
> in file gtk_dispatch.cc, line 50, the event (with its callback) is
> erased  from the map, but the callback is called a few lines later -
> the handler  gets never called because the slot object is not valid
> anymore.
>
When you move the callback invocation before the removal loop code,
does it work then?

Cheers & Thanks,
       Rotty
-- 
Andreas Rottmann         | Rotty@ICQ      | 118634484@ICQ | a.r...@gm...
http://yi.org/rotty      | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint              | DFB4 4EB4 78A4 5EEE 6219  F228 F92F CFC5 01FD 5B62

Say NO to Software Patents! -- http://petition.eurolinux.org/

Re: [Libsigcx-main] bug in GtkDispatcher::timer_callback()

[klaus t. <tri...@ao...>]

----- Original von:  Andreas Rottmann <a.r...@gm...>:
> > "Klaus Triendl" <tri...@ao...> writes:
> > hi,
> >
> > while testing timeouting handlers in GtkDispatcher i found a bug
> > (version  0.6.4):
> >
> > in file gtk_dispatch.cc, line 50, the event (with its callback) is
> > erased  from the map, but the callback is called a few lines later -
> > the handler  gets never called because the slot object is not valid
> > anymore.
> >
> When you move the callback invocation before the removal loop code, does it work then?
> 
> Cheers & Thanks,
>        Rotty

yes, it works then;
what i did was assigning the callback to a local copy which i called at the end of the method.

can you please tell me when you release a new version (debian package)? because then i would make my program dependant on that new package version.

thx!

--
http://triendl.info
klaus triendl


-------------------------------------------
Versendet durch AonWebmail (webmail.aon.at)

Re: [Libsigcx-main] bug in GtkDispatcher::timer_callback()

[Andreas R. <a.r...@gm...>]

klaus triendl <tri...@ao...> writes:

> ----- Original von:  Andreas Rottmann <a.r...@gm...>:
>> > "Klaus Triendl" <tri...@ao...> writes:
>> > hi,
>> >
>> > while testing timeouting handlers in GtkDispatcher i found a bug
>> > (version  0.6.4):
>> >
>> > in file gtk_dispatch.cc, line 50, the event (with its callback) is
>> > erased  from the map, but the callback is called a few lines later -
>> > the handler  gets never called because the slot object is not valid
>> > anymore.
>> >
>> When you move the callback invocation before the removal loop code, does it work then?
>> 
>> Cheers & Thanks,
>>        Rotty
>
> yes, it works then; what i did was assigning the callback to a local
> copy which i called at the end of the method.
>
> can you please tell me when you release a new version (debian
> package)? because then i would make my program dependant on that new
> package version.
>
You can subscribe the PTS for libsigcx, see
http://packages.qa.debian.org/libs/libsigcx.html. I think you should
get email notification with new uploads then.

Regards, Rotty
-- 
Andreas Rottmann         | Rotty@ICQ      | 118634484@ICQ | a.r...@gm...
http://yi.org/rotty      | GnuPG Key: http://yi.org/rotty/gpg.asc
Fingerprint              | DFB4 4EB4 78A4 5EEE 6219  F228 F92F CFC5 01FD 5B62

Software Patents: Where do you want to stifle inovation today?