IBMTTS on slackware

Fri Jun 6 15:36:27 EDT 2008

Hello,
I think (going from memory and not looking back at messages on the orca
list) some of the problems regarding speech-dispatcher and orca come
from is that the orca team and speech-dispatcher teams have slightly
different oppinions on what should do what. Gnome-speech is the main
supported output for speech in orca, I think the speech-dispatcher one
isn't officially supported.

As far as getting gnome-speech to use alsa, you will need to edit a file
(for viavoice it
is /usr/lib/bonobo/severs/GNOME_Speech_SynthesisDriver_Viavoice.server)
and edit the location attribute (I can't remember which XML tag it
belongs to, but that element also has an attribute type="exe") and
modify the value to include aoss (eg.
aoss /usr/bin/viavoice-synthesis-driver). NOTE: this also works for
espeak in gnome-speech.

Michael Whapples
On Fri, 2008-06-06 at 00:26 -0400, Nick Stockton wrote:
> Funny that you should menchen that because I've noticed speech-dispatcher 
> spelling stuff sometimes.
> I still use it though because on my old desktop gnome-speech has alot more 
> lag than speech-dispatcher and I couldnt get the IBMTTS gnome-speech server 
> to use alsa.
> I was thinking that maybe I hadn't played around with it enough as I had 
> just set it up to try gnome and see what it was like as I use speakup for 
> most stuff apart from web browsing, audio editing and voice chat.
> ----- Original Message ----- 
> From: "Michael Whapples" <mwhapples at aim.com>
> To: "Speakup is a screen review system for Linux." <speakup at braille.uwo.ca>
> Sent: Wednesday, June 04, 2008 5:33 PM
> Subject: Re: IBMTTS on slackware
> 
> 
> 
> > Well your experience seems slightly diffeent with the voxin people, I
> > got a version of it sent to me which had the encrypted partition removed
> > when I had contacted them with the problem of the pass phrase not
> > working.
> >
> > Yes it would have been nice to have had the packages up to date, but as
> > I am using slackware I expected to need to compile the supporting
> > applications anyway for either TTSynth or voxin. There are a couple of
> > things I am noticing between the different drivers. the
> > speech-dispatcher one seems to very quickly decide to spell things,
> > where as the gnome-speech one seems to work a lot more naturally (eg. I
> > have the list set to digest mode and it comes from
> > speakup-request at braille.uwo.ca, with speech-dispatcher it spells out the
> > request word, where as gnome-speech says each word). Thinking about
> > gnome-speech, when I recompiled it to include IBMTTS I noticed it also
> > has a driver for eloquence. Why has viavoice become most talked about on
> > linux, and which is most up to date with current libraries (as up to
> > date as they may be, which uses the least obsolete), what are the
> > differences, etc?
> >
> > From
> > Michael Whapples
> > On Tue, 2008-06-03 at 20:27 -0400, Nick Stockton wrote:
> >> Well the file can be extracted with debian and ubuntu so what you could 
> >> do
> >> is edit the install script and see what commands it is using to mount the
> >> image.
> >> Then you can mount that file under Ubuntu or Maybe GRML I think it should
> >> work too.
> >> then copy the stuff in the mounted dir to a tempory folder and then 
> >> umount
> >> and delete the image.
> >> After that copy the extracted files back to the place where the image had
> >> been mounted I think it's the mnt dir in the voxin directory.
> >> then edit the installer script and remove the mount and umount commands.
> >> After that I think it would then copy the files from the dir with out
> >> mounting anything and you can just make a tar ball of that package and 
> >> just
> >> use that one if you ever have to install again and junk the one you got 
> >> from
> >> voxin's site.
> >> I think ttsynth is the better package but I guess when buying cheep I got
> >> what I payed for but at least it does work on my debian system now that 
> >> I've
> >> stopped using slackware.
> >> I think slackware is the most stable of all the gnu/linux flavors but I 
> >> got
> >> tired of having to google for packages and then googling all the packages
> >> they needed to run and compiling/installing/configuring them so I use 
> >> debian
> >> now.
> >> ----- Original Message ----- 
> >> From: "Michael Whapples" <mwhapples at aim.com>
> >> To: "Speakup is a screen review system for Linux." 
> >> <speakup at braille.uwo.ca>
> >> Sent: Tuesday, June 03, 2008 6:36 PM
> >> Subject: Re: IBMTTS on slackware
> >>
> >>
> >>
> >> > On Mon, 2008-06-02 at 18:47 -0400, Nick Stockton wrote:
> >> >> By the way I'd expect less support than you get with ttsynth if you go
> >> >> with
> >> >> voxin *grin*.
> >> > I had come to the conclusion before that I couldn't get less support
> >> > than ttsynth has, as on the ttsynth site it says that no personal
> >> > support for the product will be given. So I came to the decission that
> >> > it is best to pay less and chance support than pay more (much more) and
> >> > know that there won't be personal support for it.
> >> >> Before I baut Voxin I sent an email asking if there was any 
> >> >> differences
> >> >> between ttsynth and voxin besides the name and the price and I got a
> >> >> email
> >> >> back with a single line saying I should ask on the speakup list *lol*.
> >> > One thing I think might be different between the two is that ttsynth
> >> > provides the speakup connector (by the sound of it, it provides it
> >> > directly rather than working through speech-dispatcher, is that true 
> >> > and
> >> > how does it compare to going through speech-dispatcher).
> >> >> I guess when something is $5 you shouldn't expect people to really put
> >> >> their
> >> >> hearts in to selling it to you but I don't think it would have taken 
> >> >> long
> >> >> to
> >> >> send a message back with a couple of differences between the two.
> >> >> Any how in case you want to know the main differenses between voxin 
> >> >> and
> >> >> ttsynth are
> >> >> voxin came in a tar file with an install script that installed the 
> >> >> files
> >> >> stored in a incrypted image and included debian and ubuntu .deb 
> >> >> packages
> >> >> for
> >> >> installing the speech-dispatcher module and gnome speech drivers but 
> >> >> they
> >> >> were already out of date by the time I had gotten them.
> >> >> Voxin did not come with the libs and header files from the IBMTTS SDK
> >> >> included which are needed for installing the ttsynth-say,
> >> >> spk-connect-ttsynth and the gnome speech driver so I had to download 
> >> >> and
> >> >> install them manualy.
> >> >> ttsynth comes in boath rpm and deb files, includes the files from the 
> >> >> SDK
> >> >> needed for compiling ttsynth-say spk-connect-ttsynth and the gnome 
> >> >> speech
> >> >> driver and the install files wern't incrypted so you can use alien 
> >> >> just
> >> >> to
> >> >> convert and install on slackware.
> >> > That encryption part is now getting me, I keep trying to enter the
> >> > passphrase and it keeps saying its wrong. I am sure slackware is
> >> > providing all the encryption stuff (cryptoloop as a module and aes
> >> > compiled in (although I have recompiled a kernel with it as a module as
> >> > well)) and I have tried installing it in GRML and get the same. I have
> >> > contacted oralux for support on this, lets see what my response is.
> >> >> *grin* that did not take long to write at all.
> >> >> I went ahead and got voxin anyway as I guessed that it would be the 
> >> >> same
> >> >> product rebranded and thought it would install better beeing in a tar
> >> >> ball
> >> >> rather than ttsynth's rpm and deb packages.
> >> >> I was quite rong how ever I didn't know that the install files were
> >> >> stored
> >> >> in an incrypted image that was mounted using the install script and I
> >> >> didn't
> >> >> know that it would be missing the SDK which I think should have been
> >> >> included in with the voxin package instead of a bunch of outdated 
> >> >> binarys
> >> >> stored in debian packages.
> >> >> I was able to make voxin work with slackware after a while but I'd say
> >> >> that
> >> >> ttsynth had the better packages and I should have payed the extra $35 
> >> >> to
> >> >> get
> >> >> it as it would have saved me lots of trouble.
> >> > May be I should have followed your advice, but I made the same
> >> > conclusions you had (with the extra one that surely the encryption 
> >> > won't
> >> > be a problem) and I spent my $5 (actually 4.29 euro) on voxin. Could 
> >> > you
> >> > enlighten me on the encryption problem?
> >> >
> >> >
> >> > _______________________________________________
> >> > Speakup mailing list
> >> > Speakup at braille.uwo.ca
> >> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> >> >
> >> > __________ Information from ESET Smart Security, version of virus
> >> > signature database 3156 (20080603) __________
> >> >
> >> > The message was checked by ESET Smart Security.
> >> >
> >> > http://www.eset.com
> >> >
> >> >
> >>
> >>
> >>
> >
> > _______________________________________________
> > Speakup mailing list
> > Speakup at braille.uwo.ca
> > http://speech.braille.uwo.ca/mailman/listinfo/speakup
> >
> > __________ Information from ESET Smart Security, version of virus 
> > signature database 3156 (20080603) __________
> >
> > The message was checked by ESET Smart Security.
> >
> > http://www.eset.com
> >
> > 
> 
> 
> 