[gta02-core] To focus, or to re-focus ?

[Werner Almesberger]

?lvaro Lopes wrote:
> Why did we lag so much ?
> 
> Well, I see several reasons. Let me enumerate those:

Hmm, I don't think the market perspective should have a major impact
on gta02-core. It was never the goal of gta02-core to make competitive
hardware, but rather to acquire and validate the skills and tools to
do so.

It would be foolish to try to mass-produce the gta02-core device. But
if one wanted to, one could make a considerably more up to date unit
with relatively minor modifications, and reusing the process used for
making gta02-core. Note that this would still need a fair amount of
money, for prototype runs, field tests, certification, and all this.

In fact, I consider it a feature that gta02-core isn't meant to be a
"real" product, because it would be extremely difficult to defend the
effort spent on evaluating, fixing, and making tools if we had real
investors breathing down our necks. (Of course, a little help from a
philanthropic multimillionaire who doesn't mind if the ROI takes a
bit longer - if it comes at all - would be nice to have :)


For me, our main weakness are the components. For a number of
reasons, Openmoko Inc. didn't have any control over the components
for a very long time, and it wasn't even clear if this situation
would improve.

So by the time we had the schematics largely done, have found an SMT
line, and have figured out the key parameters for making the PCBs, we
couldn't really move forward with major items (i.e., the layout),
without knowing the component situation.

Another issue is that I underestimated the amount of cheerleading the
project needs (or, to put it differently, that it hadn't required the
critical mass where a cheerleader isn't necessary), and the
interruptions I would get from real life interferences, such as
moving and remodeling of my new apartment (still going on, but
mercifully coming to an end soon). So whenever something else kept me
away for a few weeks, work would slow down and eventually stop.

These stoppages caused also smaller bits of work to be unfinished,
such as the remaining open issues in ECNs, or reviews of the BOM.
None of them are truly showstoppers, but each such problem makes it a
bit harder to move forward.


Regarding the market, I think Android is considerably more of a
competitor than the iPhone. Many parties are interested in having an
open platform, and while Apple make it easy to compete with them on
openness, it's harder to convey the benefits of being more open than
Android.

However, we can avoid this issue by considering this kind of openness
as a mere enabler (poof, Apple vanishes :-) and putting emphasis on
the open hardware. This appeals to a different clientele, but they've
been interested in the Openmoko project for a long time already, so
it's not really something new. Also, opening the hardware design is a
key element of gta02-core already for engineering reasons, so nothing
changes there either.


To be continued ... I'll be beset by construction workers in a bit,
so I have to get ready to face the day.

- Werner