Re: audio/flite: remove flush in play_wave()

[adr <adr%sdf.org@localhost>]

On Wed, 21 Feb 2024, Brian Buhrow wrote:

> but I wonder what you mean when you say eflite doesn't work with the newer
> version of Flite?  Is the actual interface different, or does Flite behave in a manner that
> eflite doesn't expect?

Your use of flush instead of drain (the way around in flite
nomenclature) makes the code you use responsive, you delete several
letters in yars for example, and the speech stops inmediatly and
the next letter is spoken.  That change makes new flite flush (in
the sun meaning) the buffer before it starts to be sent to the
audio device and then some other parts are sent, others not. Without
the patch, defining correctly audio_drain_sun() and audio_flush_sun()
(I mean, according with flite's meaning of these concepts) eflite
process stops at some point, the text-to-speech synthesis ends
prematurely. This is just testing, to generate some ideas. It makes
sense to try the latest eflite release (or the github development
version) with the last flite. With the last version, after correcting
drain and flush the syntesis is complete, but you cannot stop the
text responsively.  I know this sounds all crazy, I told you, the
fact that your holy trinity yars-eflite-flite works is the result
of a planetary conjuction.  Of course I'm joking, I've some ideas
but I don't want to confuse people even more. I'll post to the list
when I've some time to make other tests.

> Anyway, I hope these observations are helpful.

Of course they are.

Regards,
adr.