Re: mpegplayer updates

> Hi,
>
> I was looking at the mpegplayer plugin and was wondering if it is
> updated against current libmpeg2 SVN.
> AFAICS it isn't, so I searched a bit what changed and came up with
> these changes:

libmpeg2 has been modified to have it work without malloc, decode only Y for
greyscale targets, unneeded bits removed like checking the CPU type at
runtime and optimized for particular CPUs. It won't be a simple
conflict-free, drop-in update.

Michael

Re: mpegplayer updates

Michael Sevakis wrote:
> libmpeg2 has been modified to have it work without malloc, decode only Y for
> greyscale targets, unneeded bits removed like checking the CPU type at
> runtime and optimized for particular CPUs. It won't be a simple
> conflict-free, drop-in update.

I don't think he meant that. He found that they have fixed a few bugs 
that may still be there in our version, and that we should probably fix 
them in our end as well.

Linus

Re: mpegplayer updates

> Michael Sevakis wrote:
> > libmpeg2 has been modified to have it work without malloc, decode only Y
for
> > greyscale targets, unneeded bits removed like checking the CPU type at
> > runtime and optimized for particular CPUs. It won't be a simple
> > conflict-free, drop-in update.
>
> I don't think he meant that. He found that they have fixed a few bugs
> that may still be there in our version, and that we should probably fix
> them in our end as well.
>
> Linus

I didn't actually check the links and should have said "Might not..."
instead of "It won't...". I guess I'll do that while I'm at my email station
since I'm curious.

Michael

Re: mpegplayer updates

>
> I didn't actually check the links and should have said "Might not..."
> instead of "It won't...". I guess I'll do that while I'm at my email
station
> since I'm curious.
>
> Michael
>
>

I guess I can toss that stuff in there. Most of it isn't even relevant to
our modded, one-processor-type version.

Michael

suggestion for the steering board

Re: suggestion for the steering board

2008/7/2 alex wallis <alexwallis646 <at> googlemail.com>:
> I think that possibly members of the steering board should not be allowed to
> sit twice on the board in consecutive years,
Why not? if they are doing a good job leave them be.

> especially in rockbox, also if the members of the board were rotated this
> would mean that there were always fresh ideas coming in, and also it would
> avoid the board becoming a kind of dictatorship
you've misunderstood the point of the board... its not possible for it
to be a dictatorship or stlal new ideas... its whole purpose is to
solve disputes, new ideas come from new devs like always.

Jonathan

Re: suggestion for the steering board

Honestly, a forced rotation yearly might reduce consistency, rather than 
improve it. Because there's no guarantee everyone wants to be on the 
steering board, there may be a relatively small number of developers who 
are both willing to be on it, and meet the wants of the developers as a 
whole with a decent degree of confidence. If you force the developers to 
be unable to vote for their top 5 picks, all you're doing is saying 
"Every other year you can only have the second-best steering board 
instead of what you feel is the best."

If a board is bad, it will naturally not be chosen again. If it's good, 
there's not a real problem with members of it continuing the next year.

Re: suggestion for the steering board

On Tue, Jul 01, 2008 at 03:30:29PM +0100, alex wallis wrote:
> Hi list.
> I was just reading about the rb steering board on the wiki even though
> i new about it already, and one issue that hasn't been addressed, and
> yet i think an important one, is that of rotation of members.
> 
> I think that possibly members of the steering board should not be allowed
> to sit twice on the board in consecutive years, as 2 years is a very long
> time especially in rockbox, also if the members of the board were rotated
> this would mean that there were always fresh ideas coming in, and also it

As far as I see it, the board is not really about "fresh ideas coming in",
it's (among other things) about keeping bad ideas out, and I'm not convinced
that rotating changes anything there.

> would avoid the board becoming a kind of dictatorship where you have the
> same members for 5 years continuously. Obviously that situation would not
> be healthy. The americans change there president at a maximum of 8 years

Great example :) Does it help them any ?
Actually there is a real important difference : if you don't like it,
you can go away while actually keeping the software. It has happened
before in free software : if the leader gets in the way can't be removed,
everyone else will leave and start a new project. The other difference
is that the RSB is not a "leader" as such.

> for that reason. So I think a person should be able to sit on the board
> an unlimited number of times, as long as it is not consecutively. but
> there must be rotation to avoid stagnation.

I disagree. If someone does a good job, let him/here keep doing it. If
not, and enough people agree with you, there is this thing called
"voting"

Frank

--

-- 
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it." - Brian W. Kernighan

Creating SAPI5 voice

Re: Creating SAPI5 voice

Tomer Shalev schrieb:
> I would like to create an Hebrew TTS voice to be used for files and 
> directory naming.
> There is already exist a free Hebrew TTS engine at 
> http://sourceforge.net/projects/hebtts/
> 
> I would like to know how to make it work for the Rockbox utility.
> 
> I guess I should do the following:
> 1. Create and install SAPI5 voice (how?!)
> 2. Use the SAPI5 in Windows with the Rockbox application to create the 
> talk files.
> 
> 
> Please confirm my assumption...
> Any help would be appreciated.
> 

Hi,

i just took a short look at the website of this TTS engine. It seems 
this is not a SAPI5 TTS Engine.
But it seems it can be used from a command line utility.

So to use this you would have to modify and recompile rbutils Source code.
(In tts.cpp, add the engine name and the correct template for this Engine)

greets,
Domonoky