Stage of GTA03 development?
steve at openmoko.com
Thu Dec 25 23:27:17 CET 2008
Thanks Dr. H.N. you put it all very well.
I'll inline some additions.
Dr. H. Nikolaus Schaller wrote:
> sorry I was a bit too ironical in my answer. So let me explain.
> Am 17.12.2008 um 15:24 schrieb Leonti Bielski:
>> Nikolaus, I don't get what's wrong with my questions?
> Nothing is wrong with the questions. Only expecting precise answers
> from the project team before there is an official annoucement of a new
> device. That is what I allude to with my comments.
> I have no inside information but I know from other such projects that
> there is not really a decision being taken at a certain time that can
> be published before the first samples come out of the factory.
> It is always possible in such projects to have late changes and revise
> some previously done decisions. E.g. making the device a little
> larger. Removing a camera because the supplier can't deliver etc.
> For the GTA01 and 02 the OM hardware team has been more open - but
> they also had to discuss endlessly about changed decisions and
> slipping time schedules. This draws a lot of attention from the real
> project and does not help to make it faster.
There is sort of a rule I've tried to get people to adopt and that rule
is " are you certain?" that is, I don't think we should say things until
we are relatively ( wiggle room) certain about the details of the
product. Acting otherwise just raises expectations that are universally
dashed. It's an osborne effect squared. There are critical milestones
That everyone who followed the launch of Freerunner should remember.
I showed the phone at CES jan 07. we were just entering DVT. design
verification test. DVT, you can HOPE, will last about 2 months for a
given product. Then comes PVT, production verification test, likewise
lasting two months.. YOU CAN HOPE. finally comes MP, mass production.
Again, you can HOPE this is about two months from start to first phones
Month X start DVT
month X+2 start PVT
Month X+4 start MP
Month X+6 FCS ( first customer ship)
Month X+6.5 Shelf date.
That is a NOMINAL schedule.
During any phase of this schedule design changes both major and minor
can occur. The probability of changes goes down from month X. ordinarily
I start to feel comfortable around month X. I'm not at that point.
In addition, there are other schedule drivers.
1. State of the software. suffice it to say GTA02 shipped a wee bit
early WRT to software stability and completeness. That will change
with GTA03. With Freerunner we were really pulled in two directions,
pulled in one direction by the 'release early, release often philosophy"
and pulled in the other direction by our desire to ship something that
was more consumer ready than the Neo1973. This time around we will err
on the side of consumer ready. The "back to the basics" approach
will rule in GTA03.
2. The state of Freerunner. Not many know this but the GTA03 schedule
has taken hits so that we could support the Freerunner better. That's an
on going concern and a balancing act. In the end it is a no win
decision. That is, on one hand some will demand that each and every
Freerunner issue be settled and on the other hand some will want their
GTA03 ASAP. Sean, Wolfgang and I own that decision and the inevitable
and righteous complaints that result from it. It goes with the territory.
3. the state of the channel. One major difference this year is we have
partners with businesses, our distributors. If they don't sell, we don't
sell. That entails several things that people may not be aware of. Let
me explain. First and foremost the disty will be part of the launch
early on. they will get phones to test early in the process. Their
feedback about the stability of the software and the marketing plans
and messages will be vital. If it's not ready in their eyes, then that's
critical feedback. Secondly, the disty have to have a good freerunner
business to make for a good GTA03 business. Market development , like
product development, has its bugs ( hmm maybe a trac system is in
order...) So from a marketing perspective we will be putting additional
effort into the Freerunner to improve the strength of the channel. It's
good to remember that since Jan of 2007 we had to build a phone and a
sales channel, no small feat.
Luckily we have community support on all fronts.
Now, since this is an open source project everybody wants to be included
on every decision. Much as I would like to implement a direct democracy
WRT to every LOC, every trace on the PCB, every resistor and cap, every
subtle curve in the ID, every boss and screw in the mech design, lines
of responsibility are inevitably required to get things done. The
compromise we have chosen is one that allows people to change the
decisions we have to take. Change the code if they like, change the ID
if they like. And with the right program in place we even allow people
to change the EE design if they like. There is, I recognize, a subtle
loss of freedom in such an approach: everyone is free FROM the shackles
of closed source, but not everyone is equally free TO modify source code.
It requires skill, dedication and time. Everyone is free FROM the
shackles of closed ID, but only those with the tools, skill, and
monetary resources are truly free TO create a new ID. And lastly those
who have the tools, skill, and resources are free TO change the EE.
Someday making products may be as easy as making breakfast in the Jetsons.
But even then you might push a button and get "hot fudged pizza"
or hifi music from the garbage disposal.
>> 1. About the case design - it was a poll on ML about different types
>> of case - with keypad, without, slider, etc. So I think it make sense
>> to ask what was finally chosen.
> It could also be that nothing has been choosen yet. Or several
> variants of which one will finally reach production. So what should OM
> officially say today?
The ID has been selected.
>> 2. GSM chip:
>> Look in here how many chips were considered:
>> It makes sense to ask which one was chosen, doesn't it?
> Mickey has given a statement showing a clear preference. But until it
> has been purchased in quantities, shipped to the factory and prepared
> for production, it is not finally "choosen".
Yes. Anything can change.
>> 3. GTA03 is internal codename, like GTA01 and GTA02, but they are
>> called Neo1973 and Neo Freerunner. I seriously doubt GTA03 will go on
>> sale with GTA03 name.
> Yes, I doubt as well. But fixing the brand name is not required until
> approx. 2 months before launch. If we assume (which can be completely
> wrong) that launch is in Summer - nothing could have been choosen at
I have not come up with a name yet. It won't be GTA03. I'm taking
suggestions of course.
> As said before, I don't know about their internal project plans and
> how good they achieved any milestones.
>> 4 and 5, don't want even bother to comment this.
> There is no specific availability announced. And no price. So the only
> thing we can guess from the outside is that they will have a
> comparable price to other devices and devices will come when they are
And further there is no assurance that we won't ship something entirely
different than GTA03 before the GTA03. or ship variants of GTA02 or
variants of GTA03 before the GTA03. The best thing I can do is be a good
loss pass filter.
>>>> As proud Freerunner owner I'm interested in how the development of
>>>> GTA03 goes.
>>>> I'm not going to buy it - there is still a lot of fun with FR, but
>>>> would be interesting to know something about next open phone.
>>>> Wiki page is outdated and there is no updates on ML.
>>>> Does someone know how everything is going with gta03?
>>>> 1. What case design is chosen?
>>> A good one.
>>>> 2. What gsm chip?
>>> The best one!
>>>> 3. How will it be called?
>>>> 4. When is release date (+/- 3 month :)) ?
>>> 12:00 (I don't know on which day)
>>>> 5. Price estimate.
>>> Comparable to others.
>>> Openmoko community mailing list
>>> community at lists.openmoko.org
>> Openmoko community mailing list
>> community at lists.openmoko.org
> Openmoko community mailing list
> community at lists.openmoko.org
More information about the community