[orca-list] OT, espeak has been forked
- From: kendell clark <coffeekingms gmail com>
- To: orca-list gnome org
- Subject: [orca-list] OT, espeak has been forked
- Date: Tue, 15 Dec 2015 10:58:22 -0600
hi all
As some of you may or may not know, about a week or two ago, espeak was
forked into a new project, called espeak-ng, for next generation. Reece
dunn, the developer of espeak for android, has done this primarily
because multiple people have tried reaching jonathan duddington, the
main developer and have failed. Some have gone so far as to get his
address in the UK, and his phone number, both of which no longer work.
The conclusion is either that he is dead or no longer active, so espeak
was forked. I'm writing this mainly because it looks like the upstream
espeak from which espeak ng was forked from might not see much
development. There have already been a ton of changes done to espeak ng
that aren't in espeak, including the complete removal of espeak edit,
the gui program needed to compile the phoneme data. This has been murged
into the espeak program itself, so you can now do espeak
--compile-phonemes and espeak --compile-intonations instead of having to
use a clunky gui that has never been accessible on linux. A QT gui is
planned that will make it easier to manage this stuff but it hasn't been
written. I'm also wondering if we should switch over to this fork? And
by we I mean all the major linux distributions. Ubuntu, vinux, sonar,
arch, fedora, etc? I've already got reece dunn merging my fixes into
espeak periodically, just like he does with the italian fixes, and jamie
and mic from nvda have already switched over or are planning to.
Espeak ng should work out of the box with speech-dispatcher as far as I
know. I've not successfully built it yet, because it's still wildly in a
state of flux, but it should work. We might need to modify the espeak
speech-dispatcher module but I doubt it. Honestly that module needs an
overhall anyway, but it isn't necessary I don't believe, since the
espeak API hasn't changed much, only been added to to allow the
compilation of phonemes and intonation data. The only thing that might,
maybe, be necessary is a tweak to say espeak (next generation) or
something to differentiate it from espeak. The main differences between
espeak and espeak ng are probably not of much interest to most people,
other than that you can now do the standard ./configure, make, make
install steps to compile it, which the upstream espeak couldn't do.
THoughts?
Kendell clark
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]