[Arm-netbook] Help Taiwanese open-source GPS hardware a bit

[Paul Sokolovsky]

Hello,

On Wed, 25 Dec 2013 13:36:28 +0000
Pierre-Hugues Husson <phhusson at gmail.com> wrote:

> Excerpts from Paul Sokolovsky's message of 2013-12-23 04:41:56 +0000:
> > Hello,
> > 
> > Sorry for potential spam.
> > 
> > There's a campaign at Indiegogo to develop a board with GPS MCU
> > which would allow to run user programs alongside the GPS stack.
> > It's unclear how open the GPS stack itself would be, but even if it
> > will be a binary blob, it's still great progress comparing to
> > current GPS chipsets. As an extra, MCU feature SPARC architecture,
> > which alone worth to be present in every geek's personal computing
> > museums. And early bird price starts at pretty low for 32bit
> > "maker" scene figure of $15.
> It's 20k minimum units, so I wouldn't call it cheap, by itself.
> And I can't really see why it is better than currently existing GPS
> chips. Well, it's better (and/or worse depending on the point of
> view), because it is running on the main processor (better
> integration, easier for someone to write own GPS stack (if GPS
> peripheral documentation exists !)). Anyway, if nothing is specified,
> assumes it's closed. And current situation for GPS chips, is that
> they are using standard NMEA/AT commands, while using this GPS chip
> would make the protocol specific to this chip.

Of course, as you know, most existing GPS, and not only GPS, chips have
"main processor" inside, and all those WiFi/BT/GPS stacks run on it as
software. It's usually cost ineffective to do it otherwise, after all,
it's not a Bitcoin with its twice-SHA256 trivia and value bubble. The
only difference is that this chip has a disclosed architecture (which
is not 8051) and support to "side-load" your applications to run
alongside - as an open feature (you don't need to sign NDAs and pay few
$Ks for that). You can try to do market research and see how big is
that (I know of no such production GPS chip).

They surely won't give you GPS stack as open-source and datasheet for
GPS demod peripherals - nobody does that, so they won't either. It's
just a first small step towards commercial open GPS.

Regarding NMEA, it's a bit moot point, because first thing you do
with those NMEA textual strings is parse them into memory values,
wasting CPU cycles for that. If you're running on the same CPU, you
don't need NMEA, you need API calls to get current coordinates, time,
etc. But I'm sure there will be NMEA support either builtin, or as
an Arduino demo sketch ;-).

> 
> Still, this board has many advantages, like they say, it's a rather
> high end micro-controller, cheap (assuming they hit 30k$) compared to
> arduino, with arduino (IDE and shields) compatibility.

It appears there's single guy from SkyTraq side overseeing this
campaign, and he doesn't seem to have experience with crowdfunding, and
doesn't know that it requires full-time marketing and PR work during
campaign. They just think that if they offer great product cheap,
that's enough from their side. Nah, it doesn't work like that. If it
was on Kickstarter with a bit of PR, they could easily hit $100K-$200K
(that includes doubling prices, which would be still cheap for
Kickstarter).

Bottom line: the project is not going to reach its target at current
pace, so if you care, please spread the word. Disclaimer: I'm not
affiliated with them, but I did *basic* due diligence check (enough to
spend ~$15 for me), and the offer appears to be legitimate. It's also
"Fixed Funding" campaign - if they don't reach target, you get money
back.


-- 
Best regards,
 Paul                          mailto:pmiscml at gmail.com