Re: Proposed patch for speech problems

["padraig o'briain" <Padraig Obriain Sun COM>]

remus draica wrote:
> 
> On Wednesday 18 August 2004 13:58, padraig o'briain wrote:
> > remus draica wrote:
> > > Hi Padraig,
> > >
> > > I deleted an empty line by mistake from first patch. I also missed the
> > > error message when I checked. Sorry about that. The new patch is
> > > attached.
> > >
> > > First change in _speech/libsrs/srs-gs.c_ is fix for bug 150377.
> > > Second change in _speech/libsrs/srs-gs.c_ is part of fix for 150261.
> >
> > This change makes me nervous I do not understand the change. It looks like
> > a hack rather than a fix_at_source.
> 
> Sometimes, the START marker from gnome speech may be received by gnopernicus
> before returning from say function. This can be seen in http://
> bugzilla.gnome.org/attachment.cgi?id=30651&action=view, a patch for bug
> 150048 for the text with id = 185.
> 
> During turning off the speech (or exiting from gnopernicus, which is same
> case), this situation may occur. This wariable is set when the callback is
> received before returning from gnome-speech say (in this case the id of text
> sent to gnome-speech is unknown) and it inform gnopernicus to react to START
> marker after returning from gnome-speech say function.
> 
> The desired behaviour (and that assert was put there for it) is to _not_ have
> the callback called before the returning os gnom-speech say function.
> 
> The variable is set to FALSE next time when gnopernicus speech module is
> initialised again.
> >
> > > Change in _speech/libsrs/srs-speech.c_ is fix for 150048.
> > >
> > > First change in _speech/libsrs/srs-xml.c_ is fix for 149408.
> > > Second change in _speech/libsrs/srs-xml.c_ is for 150261
> > >
> > > Changes in _srcore/srspc.c_ are for 150261 and (I suppose) 150279.
> >
> > The bug number 150279 is wrong.
> 
> The number is 150278.

I would like to see this bug updated and marked as fixed with fix in CVS HEAD.

Is there any news of the cause of the warning messages reported by Dana?


> >
> > I would prefer a patch rather than a new tarball for safety reasons but
> > I am nervous about this patch because of the new warning messages Dana
> > reported in another email. I would also like to see bug 150261 updated. I
> > do not see that we need fix for bug 149408 in cinnabar build 17.
> >
> This bug may be relevant  desktops in other language than english. In this
> case, only a part of text will be reported.
> Do you prefer to remove it?
> 

I would prefer patches only for stopper bugs in cinnabar build 17.

> Remus
> 
> > Padraig
> >
> > > Do I need to create a new tarball as Bill has suggested? Let me know if
> > > this is necessary.
> > >
> > > Regards,
> > > Remus