On 24 Sep 2010 16:02:49 GMT
Rob wrote:

Brian Morrison wrote:
On 24 Sep 2010 09:42:02 GMT
Rob wrote:

Brian Morrison wrote:
With D-STAR unless you buy the pre-programmed DSP chips or the DV
dongle you can't legally reimplement the AMBE codec at all, and
that makes it very different from other kit that implements
unpatented modes.

The problem is that amateurs cannot live with the fact that someone
implemented a codec that is better than what they can develop
themselves.

Really? I thought that what those of us that can't live with it
thought is that we don't like technology that locks out homebrew.
That's what the use of DVSI's AMBE codec does.

Not really. You can buy the chip that implements AMBE and use it
as part of a homebrew design.

I know, but that's *exactly* what a lot of people don't want to do.

That is really not much different from buying a power transistor to
get 100 Watts of output from your homebrew transceiver instead of
developing your own, or making your own transmitter tube.

I think there is. Generating RF power from a transistor is not a
patented process, although the actual device may have some patents that
apply to it. If I had the money and skills I could build my own, but
that particular battle makes no sense. Not using a component that
includes an implementation of an algorithm that I'm not allowed to see
and understand is a different level from that, there is no secret sauce
in a power transistor but there is in the program that a DSP chip runs.

Everyone chooses their own level of components to work from. Some
build the microcontroller for their transceivers from a Z-80, an
EPROM, a CMOS RAM and some LSTTL logic. Others use an integrated
microcontroller with everything on a chip. Still others buy a small
board with a functioning computer system on it and use it as part of
their transceiver.

Indeed.


I don't know exactly why JARL chose AMBE other than because it was
the only codec available at the time. If so, they should have
thought about that a lot harder and perhaps decided to sponsor the
development of a free codec instead. That would have been really
good, but I suppose I can see that it would have introduced a
delay. D-STAR has other faults, one being that it appears not to be
extensible so that there is no way to include other codecs and
allow the correct one to be used according to the other user's set
up.

When the chose AMBE there really was no alternative. And even today,
you will not be able to find an open codec that offers speech quality
at the bitrate and bit error resilience that the AMBE codec does.

Yet. The aim of Codec2 is to provide exactly that.


It is very easy to write "then lets develop that" and apparently much
harder to actually do so.

Of course, but there are people who can do it. I happen to care enough
to encourage them and put some money into the venture to pay for their
time.


W.r.t. extensibility, it would sure be nice if codecs could be
negotiated and an alternative could be added, but it would not bring
much to D-STAR as there still would be different worlds of users that
cannot talk to eachother (those with the commercial Icom transceivers
that have AMBE and nothing else, and those with the homebrew
transceivers with open codec and no AMBE).

I'm not interested in bringing something to D-STAR, I'm interested in
bringing something to amateur radio that provides the opportunity to
break away from a proprietary solution that doesn't offer a way of
doing the self-training bit of the licence.


There would be no problem when amateurs could actually develop a
better codec than AMBE. But they have not shown they can.

Well we'll see won't we? It's taken a while to find people with the
necessary expertise but Codec2 is now moving forward with people
working on it that have that expertise.

Given enough time, it could be that something is developed that is
open and does not violate patents. But I think it will take a lot
of time and there is little chance that at the end of this development
there is still a userbase left that wants to buy and use products
based on it.

Except that without actually achieving this it won't be possible to
tell, I'd prefer to do it and then see what happens. Part of the
attraction of Codec2 is that it doesn't only apply to amateur radio,
it's something that can be used in other free software/hardware
projects such as low cost telephony for developing countries with poor
infrastructure.


Sometimes is it better to just admit that someone did a better job
than you could have done yourself, and just pay him for the work.

I have no problem with that, remember that the "free" part of free
software is referring to freedom, not money. But if someone refuses
to provide something that I can look inside and understand then I
won't use it.

It's called a choice.

That is your choice. But that does not mean that others are wrong.

Indeed. But until everyone has the choice between something they
control and something that they cannot control now, and probably never
will be able to, it is not possible to decide which solution is the
more sensible.

I'm not trying to kill D-STAR (although I would prefer that it had
never happened in its current form), but one of its problems is that it
doesn't encourage the best implementation because it comes in one form
only. If there are alternate ways to build compatible equipment then
the path to achieving the maximum performance is opened.

I don't want to see any of the other non-open digital radio standards
come into amateur radio either, I'm not actually against anything other
than the sacrifice of our ability to create our own designs without
having to use something that isn't free (as in freedom).

--

Brian Morrison