[Community] phones

[Dr. H. Nikolaus Schaller]

Hi Zack,
I have forwarded your question and my answers to our community
list because I think that many others are interested as well and maybe
want to add comments.

Am 14.09.2013 um 19:16 schrieb Zack:

> OK, I stand corrected. Three more questions:
> 
> 1. Have you looked into using Indiegogo or
> a similar crowd-funding service? Many people are finding
> success with these in firstly identifying customers and then
> honing in on what their needs specifically are. Big companies are
> pursing features like quad-core CPUs and 16 megapixels cameras
> but for a FOSS phone the market may be quite different.

Yes, we know that. Basically we did it with a self-organized
kickstarting campaign (before it was called such). And the GTA04
is the *result* of such an activity.

The key problem is not running a kickstarter/indiegogo or similar,
it is that we don't have access to an Überphone technology as
Canonical did apparently have.

So we can't offer even dual-core or 8 MPx cameras... We can
just offer what we have designed so far (without giving up some
basic principles like 99% openness, most components available
at major chip distributors).

The other problem is that nowadays a successful campaign must
be run extremely professional, because it competes with Apple,
Samsung etc. Kickstarters are no longer a method to fund some
geek projects, but have become a standard sales channel like Amazon or
Best Buy. Just a different form of payment and delivery terms.

You can see it as picking a (sometimes) refundable voucher
from a shelf and paying at the cash point.

I.e. participating needs a big marketing budget in advance,
because the expectation level did go up.

At least this is the perspective of someone selling through a
kickstarter. A pledger may still dream he is funding small and
garage projects...

For me the most prominent campaign of this category is the
Omate Truesmart. They must have a big venture capital budget to
run development, production and the kickstarter campaign
in parallel. They also mention that they have a marketing team
that is working 7/24 while the campaign runs. I.e. team is at
least 2 persons managing it full-time (and nothing else).

And, they did set the bar really low, so that the campaign was
almost immediately successful. Any maybe they would have done
it anyways. But so it also helped to strech the campaign by factor
9 in the remaining time.

I just wonder what they would have done if the strech factor
had only been 1.1 or 1.2 :)

So in summary, a product does not sell better because one
runs a kickstarter. It just becomes better known to a specific/different
audience - which may be the wrong one for the product.

Or the product may be the wrong one for the audience.

> 
> 2. As people have known for a long time, it is often
> not necessary to use fast hardware when clever use of
> software can accomplish the same. Have you considered
> focusing on cheaper hardware but better software?

Well, the original Openmoko was cheaper hardware and
was optimized as good as possible over the years and
works well. Still, people do expect higher performance
like UMTS and more memory, built-in sensors etc.

> 
> 3. Have you tried directly contacting Asian vendors
> who make cheap $50 smart phones to see whether they are
> willing to produce hardware for you that is more open?

> They have been doing interesting & difficult things e.g. putting
> phones into wrist watches, for a long time. It appears
> that Allwinner is open to FOSS.

Yes. We (this community and its antecessors) did try this for several
years before deciding to try to develop our own hardware. Basically
our core business model is NOT buying something cheap from Asia
and trying to get it opened afterwards. It must be free and open by
design. And this makes it more expensive.

The experience is that if a company is producing and selling
for 50 USD, they use some very closed chipset (e.g. MediaTek)
and don't earn any money from supporting us. And they are not
even allowed to do so by NDAs with their chip and software
suppliers or they would not get the chips at 5 USD...

So we basically have to *pay their engineers* to help us to get
a device more open. And if we put that money on top of their
production cost, the 50 USD are no longer achievable.

Mainly, the OMAP3 costs ~40 USD because it is already almost
open and well documented (~6000 pages) and is available in
single chip quantity.

BR,
Nikolaus Schaller

> 
> 
> Cheers,
> Zack
> 
>> Hi,
>> 
>> Am 14.09.2013 um 16:45 schrieb Zack:
>> 
>>> I suppose the OpenPandora project faced the same problem but
>>> they outsourced part of the work to Korea.
>> Only the plastics production. The electronics is produced also here in Germany.
>> 
>>> Have you looked into what
>>> process they used? Their quantities are similarly low.
>> No, they produce around 4000 units. And that makes a big difference.
>> 
>>> Perhaps a modular approach would work better, such as by using
>>> this type of solution:
>>> https://www.sparkfun.com/products/10138
>>> https://www.sparkfun.com/products/9533
>> Hm. We already use a similar module - and if you look at the prices
>> at sparkfun, we have very similar prices...
>> 
>> The only solution is that we produce in bigger batches, but we don't
>> get enough people interested in what we have. Most of them want to
>> have even more (e.g. Dual Core, 2 GHz, LTE) at lower price.
>> 
>> BR,
>> Nikolaus
>