[g-a-devel] Happy patch bonanza (more patch bonanza)
Willie Walker
William.Walker at Sun.COM
Sat Jul 1 17:32:15 BST 2006
Thanks for the affirmation, Enrico! I'll let this fester for more
comments for a day or so, but will probably commit your patch to
gnome-speech some time Monday morning GMT-6.
Will
On Jul 1, 2006, at 10:28 AM, Enrico Zini wrote:
> On Sat, Jul 01, 2006 at 09:25:32AM -0400, Willie Walker wrote:
>
>> Let me make sure I understand the proposal here:
>> 1) IN FESTIVAL: Rely on a convention to optionally extend the
> [...]
>> 2) IN GNOME-SPEECH: Patch the gnome-speech festival synthesis driver
>> to
> [...]
>
> Yes, that's the same as I understood.
>
>> This sounds simple enough to me. I may be misunderstanding something
>> in one of the threads on this topic, but it seems that it is implied
>> that the user will be setting the character encoding for their desktop
>> to be the same as that of their synthesis engine/voice and visa versa.
>> Should some sort of transcoding/conversion be attempted if there is a
>> detected mismatch, or is this automatically handled by the g_io
>> infrastructure?
>
> From what I understand from here:
> http://www.gtk.org/api/2.6/glib/glib-IO-Channels.html#g-io-channel-
> set-encoding
> then g_io is capable of doing recoding internally, since it says that
> "The internal encoding is always UTF-8.". I understand that the idea
> is
> that a GTK application can do everything in UTF-8 internally and g_io
> takes care of recoding when one needs to interact with the messy
> external world. Which is quite cool, actually :)
>
>> In addition, the obvious impact here is on our Telugu
>> (festival-te.sf.net) and other UTF-8 language friends - they would
>> need
>> to extend the relevant festival voices to set the "coding" parameter
>> to
>> UTF-8 and also help test this.
>
> Yes, they would need to add the coding attribute, or gnome-speech would
> try to recode their input into ISO-8859-1 (latin1), probably failing
> while trying to recode sequences that cannot be encoded in latin1.
>
> Luckily, it's very simple to add the coding attribute: see the attached
> patch for Telugu (I followed the URL you gave to download the sources).
>
>> Please let me know if I'm understanding this correctly. In addition,
>> many many thanks to both Enrico and Milan for their understanding and
>> diligence in this matter. You definitely help define what "community"
>> means.
>
> Thanks!
> /me blushes
>
>> PS - I'm out of the office for the next several days, but I will
>> release a new gnome-speech tarball for the next GNOME 2.15 deadline
>> (12-July) if we can quickly reach closure on this.
>
> That would be just fantastic.
>
>
> Ciao,
>
> Enrico
>
> --
> GPG key: 1024D/797EBFAB 2000-12-05 Enrico Zini <enrico at debian.org>
> <telugu.patch>
More information about the Ubuntu-accessibility
mailing list