On Wed, 8 Aug 2007, Onj wrote:

>> Why? Surely the *vast* majority of all users (blind, visually impaired or
>> perfectly seeing) download a voice file made by someone else?
>
> 1. I'm blind and I used too, until a few days ago, make voices successfully.

A vast majority is of course only a majority, not everyone.

> 3. I use windows XP. How can I install sigwin, and does it have to be
> sigwin?

First, this isn't even integrated in the sources or SVN yet and even when it
is that, you're fully free to work on and maintain a system outside of cygwin,
if that is if truly is what's needed for you (and I then mean you in the
greater sense, not just you as in a single person).

Just get the scripts updated to output the new format. It's not that
different.

> I'm a power user of windows but no other operating system. I have messed a
> little, with ubuntu so laugh if you like, but would it be possible to build
> the voices in that as it's a Linux varient, or am I speaking out the wrong
> end?

Yes, it is perfectly possible to build voices with Linux. The ones we provide
one on the site are made on Linux.

The Windows-generated voices could however use the fancy proprietary speech
engines that produce better speech than what the free ones on Linux does.

> perhaps I just didn't read the Rockbox page carefully enough, that this
> change was almost thrust upon us.

This is the _users_ mailing list, we don't discuss much development here so of
course we didn't publish our intensions with this move here.

This change in the voice system has been in progress and worked on since late
2006, and we've had several patch tracker entries filed and dicussed and it
has been discussed in the development team on and off for a long while. It was
not a surprise.

If you want to speak up and tell your opinion about things we develop in
Rockbox, that is perfectly fine but you should then join the development
community, on the dev list and possibly also on IRC.

> I was happy the way things worked.

Sure, but we had reach a dead-end in some senses and we needed to introduce a
way to get out of that. This was the way we came up with, that now allows us
to introduce a trackload of more phrases to have spoken, without them breaking
builds for the smaller siblings in the Rockbox family.

> I don't know what players people used, and what particular voices I can use
> now it's a totally different system, how many voices to make

I never use voice myself. I never provided any voice files before. I now
generate and provide voice files for all targets we support.

What's there to whine about?

-- 
  Daniel Stenberg -- http://www.rockbox.org/ -- http://daniel.haxx.se/