gta02-core (was Re: OM future)
Carsten Haitzler (The Rasterman)
raster at rasterman.com
Thu Feb 25 02:39:25 CET 2010
On Wed, 24 Feb 2010 11:57:39 -0800 (PST) Rafael Ignacio Zurita
<rizurita at yahoo.com> said:
> Hi, I think that gta02-core is the key (or part of).
>
> All of us have been talking about the problems with openmoko hardware
> for years. And because all of us have different ideas and feelings
> we use this list as our psychologist :-)
>
> But still if there are a lot of different ideas and feelings
> I always read the same conclusions :
>
> - Openmoko hardware has problems
> - There is no other open linux mobile phone like we want
> (open on both, software and hardware sides : without signed
> bootloader images/etc)
>
> Also, because we write those feelings all the time, we can read
> the answers from old Openmoko devels who also have differente ideas.
> But, most of them have, as well, the same conclusions :
>
> (1) Openmoko failed internally
> (2) The Openmoko development process was not good
>
> When we look for alternatives these are always the same,
> from big companies, and with the same features:
> propietary components and perhaps monopolistic. So I guess no,
> there is no open linux mobile phone like most of us want.
>
> Then, you can find possible (future) alternatives. If you read Raster words
> he says that just a big guy (company) can do a sucessful linux product,
> because it needs money and lots of people. Perhaps that would
> be just a small step toward an complete open Linux mobile phone.
> If lot of little steps are made, some day we may get what we want.
>
> But, I also feel that gta02-core is another key to fix some
> of the conclusions that ex openmoko developers did, because
> gta02-core tries to fix the process.
>
> I read from time to time the gta02-core mailing list, and irc chat log.
> Some people think that that project goes slow and because gta02-core
> won't produce the next gta03 mobile phone, then
> it is useless and a waste of time.
>
> But no, if you read well, then you will realize that it is a
> community development process, and if this project succeeds then
> the development of a new cool open linux mobile phone may
> follow similar lines of the gta02-core process documents.
>
> That would fix the (2), and then you just need a big guy,
> like Raster says, without internal problems like there
> were at (1).
>
> With (1) and (2) solved, we would have our cool linux device
> wanted, and without bad feelings to go to the psychologist ;-)
>
> Now, I have not tried to help a lot to gta02-core, and I don't
> know why there are just few people doing all the hard work.
>
> In few months gta02-core has created and reviewed the component diagrams,
> created and reviewed the schematics (using just open source software),
> found a smt line (from an university) to do the prototipes, written
> a footprint editor.
>
> I guess that they need the components kit that Sean/Openmoko inc
> promised, and help of much people from community to continue
> (check the tasks list to do).
>
> The important thing is that the gta02-core project output will be
> the development process docs and the organisation for that.
> Proper processes and organisation, plus software (linux+fso+shr(distros))
> will give us a great opportunity to find some
> company without internal problems and say them :
>
> "Hey, we know how to development an open linux mobile phone
> like it should be, so please, start a new project for
> that kind of hardware and we will point you how to do it" ;)
>
> Now, I need to check how to really help more and stop to
> look for a replacement (which does not exist).
>
> fso, shr, gta02-core, other distros : thanks!, and keep the great
> job you do.
openmoko managed to produce something at all... because someone (sean) was
lucky enough to find a pot of gold without strings attached making stuff
closed. such a pot of gold is amazingly rare - and if found is of limited size.
unless that pot of gold is used very smartly to bootstrap a self-funding
profitable venture it's all doomed to failure.
gta02-core is waiting for that pot of gold. as such it may wait years for it to
appear - if it ever does. you can design yourself silly - without the money to
buy components in quantities that make them sufficiently cheap and the ability
to fund the certifications, production and all the initial costs before even 1
thing rolls off the line that can be put in a box and shipped to someone...
you're goign to just wait.. and wait... and wait - and in the end your
gta02-core design will not be possible as components you used go out of
circulation and your design is the equivalent of using an 8 bit
microcontroller in todays phones - ie the world moved on and your wonderfully
open design is useless as the world moved on to LTE, 8-core multi-ghz soc's
with multiple gpu cores, and god knows what else and THAT is now considered a
phone - your old design is like saying "hey - i have an open 20mhz 386 here
with a hercules video card, 2mb ram and 75 baud modem! isn't that awesome -
it's OPEN!"... but you say that in a world of 3ghz quad-core processors with
multiple gb of ram, high end gpu's runing multiple screens with smooth 3d
although res... and 802.111n.
every month that gta02-core sits as a design and doesnt go into production is
another step in the above direction. sure. you may argue that "i dont need
multiple screens". and "i dont need more than 640k of ram" or "i dont need more
than 75baud" etc... but you know how silly that sounds if you said it now.
99.999% of the world can't live on that for their COMPUTER needs.
so - you now have to keep finding new soc's to design for and new modem modules
and keep re-designing gta02-core every few months to keep up with the world -
but you're doing this all "in theory" without any production. and production
takes money. who's going to pay for it? there are no small steps here. there's
a big costly step of getting the thing produced at all. it costs millions of $
just to get there. who here has that in their pocket and is willing to risk it
all in making some gta02-core's in the name of openness?
thats your problem - thats what you need to address. if you dont have your pot
of gold - you're spinning forever going nowhere becoming obsolete. the openmoko
pot of gold is gone. you're not getting that back. the way i see it is - make
your compromises and work with those who have the big pots of gold and bring
them over to your side one bit at a time. you''d be surprised how much elements
already are on your side - there are just others fighting to keep things where
they are. there are some things you will never get open - i doubt u'll ever see
a schematic or case cad design - ever. thats an "almost impossible" battle. the
easier battle that CAN be won on most fronts is in the software side - kernel,
drivers, userspace, open standards, infra, core, toolkits, etc. and the big
boys with the big pots of money are surprisingly close there. rome wasn't built
in one day. fight the battles you can win - dont fight the impossible ones.
sure - it's not as glorious. it's not as sexy. but it gets you one small step
closer to where you'd like to be. you may never get it. that's the nature of
compromises. but you can get some of it at least.
--
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler) raster at rasterman.com
More information about the community
mailing list