AGPS closed source drivers = DRM for public data

Sean Moss-Pultz sean_mosko at
Thu Dec 28 18:39:31 CET 2006

On 12/28/06 1:56 AM, "Richard Franks" <spontificus at> wrote:

> I have a bit more patience on this matter because:
> 1) I believe that the OpenMoko team would love to see a 100% Open
> Source phone that pleases everybody.
> 2) I believe that there is a lot of information which cannot be shared
> publicly with the community due to NDAs or ongoing negotiations.
> 3) I believe that the FIC team have a strong product strategy for the
> future, which again, it would be foolish for them to disclose at this
> stage.
> 4) I believe that by attempting to second-guess that which we don't
> know, then standing upon that shaky foundation and attempting to apply
> pressure to companies who have a relationship with FIC, we risk more
> harm than potential reward.
> How about we give FIC a chance to deliver on their promises first?

Thanks Richard. I really appreciate these kind, encouraging, words. This is
the first (of many) devices we are working on. If enough people don't like
something the first time around, without a doubt I'm open to changing it the
next time. But please have patience now. I know my comments are less than
they have been in the past month and coming slower. This is only because
even more of my time is being used up getting this device through

Rest assured that everyone on my team is committed to bringing you a truly
open device. Let me digress for a moment to clarify some things.

Making a mobile phone costs millions of dollars. It is absolutely necessary
for a business to produce at the very least the profit required to cover its
own future risks. Therefore, as an employee, it is my duty to provide FIC
with a product that has the potential to generate a profit without a high
degree of risk.

>From a strategic standpoint, I have chosen to focus OpenMoko on the
following two areas:

    * Connectivity -- both cellular networks and wireless networks; and
    * Location -- both antonymous and assisted GPS.

I believe these are the core features that will drive handset sales for the
next two to three years. Excelling in these areas will allowing me to return
a profit for FIC and be permitted to create new products. "Why" is not
important for this discussion. "How", is crucial.

The GPS chipset we are using is a new kind of architecture that separates
the processing algorithms from the actual hardware. This solves two
important problems that have kept GPS functionality out of most GSM mobile

    1. Size -- By moving all the processing algorithms to the existing
    processor it is possible to make a GPS chipset that is 1/6 the size of
    traditional ones. Small enough for a mobile phone.
    2. Cost -- Since the GPS chipset only outputs raw data, it does not
    require an additional dedicated ARM processor. This saves about US$10.
    Multiplying this value by nine, yields the cost savings to the customer.

Yes, it requires that a proprietary library exist in user space of OpenMoko.
But, in my opinion, I need this to have the chance to sell hundreds of
thousands of handsets to A) make a profit and B) be in a position of
strength within my industry.

I think by now I've clarified the first point. Let me further expand on this
latter one. Any chance of having a mass market Free Software phone will
require incredible positional might. Vendors price and support their
components based on volume. The promise of an innovative product alone is
not very persuasive. I can go to a vendor now, telling them about the
potential of OpenMoko and say, "allow me to GPL your driver." They will end
our negotiations before things even begin. But instead, if I'm armed with a
million handset sales under my belt, I can go to the very same vendor, and
they will say, "I'm listening."

So for now, people must make their own choice: Either you want to be 100%
pure Free Software on the Application processor (at the expense of loosing
GPS functionality) or you can chose to be 'pragmatic' and use a single piece
of non-free software to have that functionality.

I'm sure we can spend lots of time debated this in the future. You can even
do this now if you like, but it's futile. We can't change anything at this
point ;-)


Happy Holidays!


More information about the community mailing list