More changes at Openmoko

Posted Jun 3, 2009 17:48 UTC (Wed) by BrucePerens (guest, #2510)
Parent article: More changes at Openmoko

Sigh. That's pretty much the end unless the ex employees can really come up with something that can be sent to a PC board shop on their own.

to post comments

More changes at Openmoko

Posted Jun 3, 2009 19:40 UTC (Wed) by laf0rge (subscriber, #6469) [Link] (5 responses)

well, I would assume that gta02-core will produce working gerber files that can be put to a pcb manufacturer and SMT factory for actual component placement.

However, it is still a 2G or 2.5G device, based on an aged SoC... many years behind the kind of hardware that proprietary competitors are shipping.

So yes, this way you can continue to design and build open hardware cellphones - but you can certainly not produce something that the typical user will get excited about.

Using a more recent SoC like the s3c64xx (intended for GTA03) or an OMAP3 or i.MX51 or whatever else you might imagine is one aspect - and it is the easy aspect. The hard part is the actual 3G/3.5G modem. Unless you go into quantities of at least hundreds of thousands, there is no business case for any modem chip vendor to help you. And then those chip vendors are not only secretive/proprietary with regard to their hardware and firmware, they also are very secretive on the external interface/protocol.

An all those electrical design issues don't even speak about issues such as mass production, QA, mechanical engineering, distribution/sales, etc.

More changes at Openmoko

Posted Jun 3, 2009 20:55 UTC (Wed) by farnz (subscriber, #17727) [Link] (3 responses)

And it faces a big problem due to the lack of 3G support - there are people like me who would like a Free phone, and would pay for it, but are provisioned with USIMs that require a 3G capable handset. I thus cannot buy a 2.5G only handset, as my network subscription simply will not function.

More changes at Openmoko

Posted Jun 3, 2009 22:09 UTC (Wed) by drag (guest, #31333) [Link] (2 responses)

This is quite possibly what OpenMoko should of done from the beginning:

http://www.gizmoforyou.com/comment.php?comment.news.40

It's the flow phone.

Instead of designing a new board from scratch they took a off-the-shelf solution from Gumstix for the main processor board.

They designed a daughter board to go with it that plugs into the main board which provides break out for mounting other devices...

So what features you want depend on how much you want to spend on the thing.

Camera, regular GSM, 3G module, etc etc. If you want 3G rather then a regular wifi phone then thats probably going to cost you another 50-100 bucks or whatever.

Very very cool.

Now I have not seen anything showing a actually working phone yet... but it is still cool.

http://www.gizmoforyou.com/comment.php?comment.news.40

Gumstix vs. gta02-core

Posted Jun 4, 2009 10:12 UTC (Thu) by pboddie (guest, #50784) [Link] (1 responses)

The idea of using Gumstix has occurred to people before: there's the TuxPhone (http://www.opencellphone.org/) and the Homebrew Mobile Phone Club has a list of similar projects (http://wiki.hbmobile.org/index.php?title=Hardware_Projects). Moreover, the availability of GSM modules like those of Telit has changed the landscape somewhat.

Although the Gizmo Flow is a good idea, I'm not yet convinced that the Gumstix Overo boards are ready for prime time. From what I've read, the kernel build for OMAP doesn't support power management properly (http://www.gumstix.net/Hardware/cat/Benchmarks-power-temp...), leaving the components running very hot. And for those wanting to "remix" their hardware, Gumstix mainboards are not openly documented, so there's obviously a gap in the market there.

Gumstix vs. gta02-core

Posted Jun 5, 2009 0:41 UTC (Fri) by drag (guest, #31333) [Link]

Cool. Thanks for the info.

Still, even though the Ovaro is not open source I think that this is probably the best approach right now for a Android phone with the combination of good performance and openness.

More changes at Openmoko

Posted Jun 4, 2009 10:03 UTC (Thu) by tzafrir (subscriber, #11501) [Link]

software components are reusable. So you don't have to rewrite the whole Linux kernel just because we have a new generation of wifi devices.

How much can be recycled out of that old design?

More changes at Openmoko

Posted Jun 4, 2009 6:12 UTC (Thu) by tajyrink (subscriber, #2750) [Link]

It's not just ex-employees, there are thousands of Neo owners who are also participating in various efforts. And you don't need to be an owner to participate in designing new hardware, only if you are participating in the various distribution/software developments. And the current A7 version of FreeRunner will continue to ship.

Of course in the long run newer hardware would be needed, but at least now for the first time I actually _start_ recommending Neo FreeRunner to people I've so far told to wait for the buzz-fixed version, which is now shipping.

And I'll happily use my FreeRunner for as long as it lasts, if it's truly so that nothing better comes along, like it looks currently.