Re: [orca-list] Ibmtts and Orca.

[Jan Buchal <buchal brailcom org>]

> > > > > "TC" == Tomas Cerha <cerha brailcom org> writes:

    TC> Willie Walker wrote:
    >> If you are unable to do the testing, I kindly request that you
    >> get Tomas to take a look at this.


Hi again,

how I said before the problem is not from my point of view in speech
dispatcher backend only. And now I can tell more. If I delete .gconf*
directory and again set the orca all works how it has, that means,
cancelling works, cursor navigation etc. 

So the question is how depends some changes in .gconf with cancelling or
not cancelling and with other problems with orca. Caret browsing in
firefox for example not works too and after .gconf* removing works
again. My feel is that configuration interface in orca has not solved
problems.

Please note that bugs we can find anywhere, I don't know a software
without it, maybe qmail only :-)

Have a nice day



    TC> Hello Will and all,

    TC> I'll take a look at it. For now, I can just say, that I have
    TC> already noticed this strange behavior. If you remember, Will,
    TC> this was the problem which I reported to you when you first
    TC> included SD backend in Orca. I simply never got the stop
    TC> commands which you said are always executed by orca on user
    TC> interaction etc. But than, with a later Orca version, the
    TC> problem disappeared and it started to work as expected. Then I
    TC> was one nice day doing something on the SD backend and updated
    TC> Orca from SVN and the problem ocured again. Since it didn't
    TC> prevent me from doing what I wanted, I didn't investigate it and
    TC> when I got back to Orca a few days later, I updated from SVN and
    TC> it disapeared again.

    TC> So what I want to say is, that this problem sometimes appears
    TC> without any change in the SD backend itself, but must depend on
    TC> some changes in Orca. I definitely don't say it is the fault of
    TC> Orca. In fact believe it there must be some problem in my code
    TC> (in SD backend itself), but I only describe what I experienced.
    TC> When we were first investigating this, you asked me, whether it
    TC> is possible that something in the SD backend is behaving
    TC> synchronously. Could you please be more explicit what you mean
    TC> by it?

    TC> In any case I updated Orca from SVN today and the problem is
    TC> here again, so at least I can investigate it. I'd be glad to
    TC> test that with gnome-speech but the Festival driver doesn't work
    TC> for me and I don't seem to have espeak support in my gnome
    TC> speech installation. Is there an easy way to get it? I'm running
    TC> Debian unstable.

    TC> BTW, here is how test-speech ends on my box:

    TC> cerha bob:~/work/orca/trunk> test-speech 1:
    TC> OAFIID:GNOME_Speech_SynthesisDriver_Festival:proto0.3

    TC> Select a server: 1 Attempting to activate
    TC> OAFIID:GNOME_Speech_SynthesisDriver_Festival:proto0.3. Driver
    TC> name: Festival GNOME Speech Driver Driver version: 0.3
    TC> Synthesizer name: Festival Speech Synthesis System Synthesizer
    TC> Version: 1.4.3

    TC> Enter desired gender ('m' or 'f'): m

    TC> Enter desired locale, or 'all' to display all voices: all 1.
    TC> kal_diphone (language english)

    TC> Select voice: 1

    TC> ** (process:9503): WARNING **: Failed to send buffer Neoprávněný
    TC> přístup do paměti (SIGSEGV)

    TC> Best regards

    TC> Tomas

-- 

Jan Buchal
Tel: (00420) 24 24 86 008
Mob: (00420) 608023021