[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

a tiny bit more on high-bit characters

I've looked just a bit further at the problems several of us are
having with characters between decimal 128 and 255.  It's turning up
in the French and German versions with bad treatment of accented
letters and in English with bad treatment of some puncuation etc
generated, for example, by emacs-w3.   

At the moment I think the problem lies within viavoice or more accurately,
 at or below  the tcl level. I looked at the strings actually sent to
 the speech-server by dtk-dispatch-queue and the characters I was
 testing were surviving unaltered to that stage, which is about when
 they leave the realm of emacs and hence emacspeak I think.  That is
 the next thing which happens to them is a process-send-string.  
I also opened the socket-listening speech server and played with
 commands like 
tts_say { this is a test }
which was fine and then
tts_say { this    is a test}
where there's an A circumflex after "this".  Viavoice says  "this A
 tilde is a test"
just as it does from emacspeak.  
I presume this is all fine on a Dectalk  as well?

The Viavoice documentation claims to use the ANSI character set but
that's the same as iso-8859-1 is it not?  
So if it *is* Viavoice the question is whether it's a fault in the
implementation or in the set-up.  
Just to make sure we're talking about the same versions, rpm -qf
/usr/lib/libibmeci50.so yields

Are there any potential issues of signed vs unsigned chars as we feed
characters with the 8th  bit set through the tcl stuff? 
Ok, hopefully more tomorrow

To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@cs.vassar.edu" with a
subject of "unsubscribe" or "help"