werner at openmoko.org
Mon Apr 13 00:17:55 CEST 2009
Gerald A wrote:
> The current/testing stuff could be
> community designed. Things that get moved to "stable/release" would
> have the OM seal of approval.
Yes, I think that would be a good model. You can often distribute
prototype development for subsystems, and the cost of entry would
be relatively small in many cases. E.g., development of the
acceleration meter subsystem could have been done completely
outside Openmoko and not connected to any specific product. You
risk not getting your prototype chosen for the real product, but
the loss would be relatively small.
Things get more difficult with those parts that you can't just buy
online, e.g., CPU, telephony, LCM, battery, etc. A highly visible
"community" project may actually have a strong enough reputation
that some companies may be interested in cooperating even if they
would normally not give you the time of the day unless you're a
Fortune 500 company.
Openmoko actually benefited from this effect: it was often given
better treatment than a company of this size could normally hope
> In this way, a couple smart hardware hackers can put together a phone with a
> built in itch-scratching device.
Ah, but now you're talking about integration. That's where you
can't distribute so well anymore. Even connecting the prototypes
can be hard. I think you would basically have to get the people
together in the same room for a week or two, along with all the
equipment needed for such a task. That way, everyone can see the
same problem when something goes wrong. No chasing of bug #1024
around the planet ;-)
I've been in projects (EU-funded research) where we did this sort
of integration, although in a less hardware-centric way. You need
a site hosting a common testbed that everyone involved can travel
to with relative ease, a rather generous travel budget, and more
time than even the pessimists will tell you you need ;-)
> Maybe this can be talked about, without too much detail?
I wouldn't be afraid of going into detail. I think that's where
Steve's and my approach differ a bit :) I often find that I only
know the right questions after having studied some part in detail,
and perhaps even having made a prototype.
In electronics, you often can't just adjust the few parameters
you don't like in a part that's great otherwise. Instead, you can
only pick a very different part where other parameters don't fit.
The tricky bit is picking the one where the parameters you don't
like don't cause major trouble.
> That depends on part price point, and device price point. Now, I'm not a
> biz guy either, but if the parts are only $0.10, and you are circular
> binning half of them, $0.20 isn't a huge impact.
Ah, but eventually you'll put those 50% bad $0.10 parts into your
$2000 prototype :-) So if one component is that badly out of whack,
you have to do careful input testing, which means coming up with a
repeatable test process, and so on. In the best case, you can manage
the risk, but you'll probably miss the resources you have to put
into it elsewhere.
> This is where things bounce over from hobby to business, IMHO. If a hobbyist
> comes up with a phone that can also function as the key for your car, that
> might be interesting.
I'd rather hope for a more conservative design - do one thing but do
it right. Otherwise, those fancy features just end up taking up too
much time. There's always the possibility to add a much larger
feature set in follow-on products.
> But, it also compounds your support issues,
> because now you guys aren't able to get community software support in
> critical areas where you would be leveraging the real power of the open
> source community. (Let's not even talk about field upgrades of firmware and
> the like).
I think early access to GTA02 wasn't done properly. We should have
involved external testers and perhaps even developers earlier. The
lack of testing then led to surprises like #1024, uSD vs. GPS, buzz,
In GTA01, we did in fact try to do this testing right and sent out
quite a number of freebies. But since the software wasn't nearly
ready for any use, only very few people could actually do something
with the device, and as far as I recall, there was no feedback.
Of course, in a way, GTA01 was the early access to GTA02. But the
software situation didn't really improve until later.
More information about the Gta03