[Gta04-owner] Invitation to indicate interest in next batch ofGTA04A5 boards

[Xavi Drudis Ferran]

On Tue 30/04/13 10:03 , "Dr. H. Nikolaus Schaller" <hns at goldelico.com> wrote:

> Another thing to think about is what kind of device (i.e. feature set) we
> need to offer next year so that 90% of the FOSDEM users are willing to
> pledge for 300 EUR. Maybe the same features as a Galaxy S4...
>

I really don't know. But I suspect that offering the same feature set that
someone else with deeper pockets won't help much. 

Somehow I think the most interesting thing would be customizability. 
I'm not good at hardware, I have the impression that the board already has 
much, only limited by size. So anything that makes it easier to use might help.
Maybe something like bringing connectors to the outside, thinking of ways 
to use extension ports with conenctors that can stand many plugging and unplugging, 
I don't know. 
 
> Recently it became clear to me how community developed software and
> hardware are inherently different compared to their non-community
> development counterpart.
>

Software can be replicated at 0 costs, hardware replication costs never
reaches 0 and only approaches it a little when making great many copies. 
On the other hand hardware prototypes are more expensive to produce than
software prototypes. This changes costs of experimentation, therefore
participation. 

Physical constraints apply to hardware and not to software, so reuse and 
experimentation are also more restricted. But I think this isn't the main factor
as electronics are relatively standarized. 
 
> Software can be developed easily by a community through being open
> source. And since distribution of open software is cheaper than for
> closed, it has even an immediate financial benefit over closed software.
>

Software can be reused very cheaply. This lowers costs a lot. 
Software evolves, hardware not so much (only by producing new units, not
replacing old artifacts). Hardware designs can be reused, but boards
can't. So sharing software is an immediate saving. You get more people
using your software, and potentially decrease maintenance costs and
leverage network effects. Sharing hardware is limited to just blueprints, 
you can't really share the hardware product per se. It's physical so only 
one person can hold it at a time. 

So the economics are very different, which does not mean that it can't be done
collaboratively. For collaborative hardware development to prosper what
we'd ideally need would be components stability, flexible production factories
abundance and low cost, and somehow less regulatory burdens (certification
and such, but I also believe regulation is useful for consumers, environment, etc.)

Open hardware will never be the same as free software, but in an ideal world
it wouldn't be so hard as with the current disarray of production (oligopolies, 
artificial obsolence, secrets, patents...).
 
> For community developed hardware, development by volunteers is also
> cheaper. But production and distribution is always more expensive,
> because
> of lower quantities.
>

It doesn't have to be always, but currently, yes. The economy would have to 
change a lot for this to be different. 
 
> In the outcome free&open software is cheaper than closed and open
> hardware is inherently more expensive. And therefore no $BRAND is
> interested in opening the hardware - because it costs additional money
> (starting with writing good documentation).
>

It costs more, but it also produces more useful products, which helps 
customers buy less stuff. So companies won't do it as long as they can
keep everyone else from successfully doing it. 
 
> So unless this mechanics is better understood by society and free&open
> enthusiasts are accepting that open hardware is necessarily more
> expensive
> than closed, there will be no big chance for open hardware.
>

On the other hand closed hardware is evolving into so tightly controlled
closed gardens that reasons for open hardware increase. It's no longer
a matter of curiosity or potential adaptability, it's privacy, freedom, 
and uncrippled functionality. At least potentially. 
 
> So we can't draw the argument that free&open hardware is cheaper.
>

Right. 
 
> Now you may ask why Arduino, Raspberry Pi, BeagleBoneBlack etc.
> are so successful.
> 
> There is one difference: they are already cheap enough to scrap them if
> they don't fulfill the requirements or if the student has learned
> something.
>

May help. But I think it's more about they're building blocks. A phone is a
computer which is a general machine, and in that sense is also a building
block but Arduino, RPI, BleagleBone... are more physical building blocks.

What I mean is that people who are willing to build new things take a lot 
of advantage of somethin glike arduino. It's low cost but it's also able to
drive very different and useful machines by adding it more or less cheap
hardware. Cheap + useful + requiring work = learning experience. 

A phone purpose is not to build things with it, just to use it for what t is for. 
People don't expect to increase its usefulness from what is shipped
(other than by software)

> And some of these projects are part of the marketing strategy of the chip
> vendor. They are interested to push their CPU to as many potential
> *developers* as possible. So it is their "CPU Evaluation board".
>

Yes. Hardware requires factories and software doesn't. So capitalism 
applies to hardware. 
 
> While our kind of devices can not get into high quantities by aiming at
> developers, learners and DIY enthusiasts only.

Not that high quantities, but might get more demand if somehow it was
very easy to use for DIY projects. I don't really know how : 
- arduino compatible expansion slots ? 
- simply more general purpose I/O in alternate case ? 
- some specification for some form of expandability, like an alternative case
with hole for cables and anchors or something for peripherals to hook into   
- Periperals ? I don't know, and interface to an eink screen, or a EOMA68
compliant gadget that you could connect to gta04 to plug it as the CPU for 
a EOMA68  tablet, laptop or console ? or a peripheral to plug video/audio equipment
so you can record from micros or musical instruments ? 

I don't know. Don't take this mail very seriously. I don't see any real solution 
that looks workable and attracts really many people.