[Accessibility] TTS API interface description + reqs update

[Gary Cramblitt]

On Monday 01 May 2006 08:40, Hynek Hanke wrote:
> Gary Cramblitt writes:
> > I have some spelling and wording changes I will post in a separate
> > message.
>
> Perhaps you can commit them in CVS directly unless it is something
> controversial.

Done.

While making the changes, I had questions about the following:

2.1
"When a string value can contain characters not present in the ASCII
table (namely various language specific characters), the data type
used in this interface is @code{wchar_t} and the corresponding
expected format for Unicode is UTF-32. When a string is marked as
type @code{char*}, it can't contain any wide characters. Only when the
implementation of the interface (for example TCP protocol) does not
allow for different types, UTF-8 encoding and wide characters is to be
used for values marked as @code{wchar_t} in this definition."

I don't understand this paragraph, especially the last sentence.  Is this 
saying that when UTF-32 cannot be used, UTF-8 may be used?

2.6.2
"In case no voice matches the given language, the synthesizer should
pick the general default voice (if applicable) or choose an
arbitrary voice. This alone is not considered an error and must not be
a reason for the synthesizer to refuse further synthesis requests
unless for some other related reason (as for example the voice being
unable to handle the given UTF-8 character range)."

UTF-8 => UTF-32 ?

-- 
Gary Cramblitt (aka PhantomsDad)