Continuing on that road, I 'd prefer to purchase 3 upgrade motherboards to fit in half the number of N900's I own for 1000 Euro total than to purchase yet another full N900 for 700 Euro's.

How about a proper expansion interface somewhere under back cover, to allow "The Original Half" arbitrary hw expansions by exploiting a mugen back cover?

Sorry I don't get the math sorted in your post, but anyway be assured that you will be able to buy bare board to retrofit into N900 case+display+otherstuff, just like original GTA04 right now does with Neo Freerunner (GTA02). This will probably save you ~100..150EUR of the price for a full Neo900

/j

Will it do USB host mode ootb?

About fre(e)mantle, is it going to be based off debian or an exact (or modified to change the closed blobs) img of the current fremantle?

Yes, see post #1 please
We got "grace of late birth" benefit here - we don't need to do all sorts of weird things like new USB PHY, to avoid a SiErr in TWL4030 chip. Meanwhile those chips are available in a fixed version, so USB shall work like it originally been supposed to.

Very exciting!!!

I've read all the thread thus far...I do like this project.. here's my thoughts..

Effectively, it's and open HW/SW mobile computing platform...

HW

I think you guys should bring in another open ARM based hardware group to get your runs up... No offense to openmoko guys.. it's a hard ask on both LEGACY communities
Getting runs up HAS to be a high priority.

Talk to similar projects/groups
IE: Single-board PC's
http://en.wikipedia.org/wiki/Single-board_computer

Or start a new one?
The board would be great for this... the N900 is 10x better then any of these single boards.
A sim slot can be useful here, for home automation/remote internet/security etc.


Also, ouya/amazon/madcatz/etc and other android gaming devices are hot.. maybe there's interest in using this board as a gaming/multimedia device?

Maybe make a run with no sim/camera, different SKU??

Does it have to ARM? A lot of SOC's on the market... not sure I buy the 'It's a closed architecture' line for all but ARM SOC's.
Look at Samsung... A lot of cheap chines tablets using there SOC's.. has to be resellers

I can't see FM/IR being delivered.. can add-on modules be used in mobiles?
I don't think it's core and people asking IMHO don't appreciate using mobile linux with a HW keyboard.

SW

As much as I love maemo, it shouldn't hold the HW back... no offense to maemo
Can always replicate the interface.. oh what happened to http://cordiahd.org ?

look at what other SOC's linux distro's are targeting... and align with that.
IE: Tablets/phones/single-boards etc

I would rather it be based of fedora/ubuntu/debian/jolla/etc and use the parent's repo's.

I think it's a great idea to use the N900 as a case.
Ideally, I would like today's HW(SOC) with today's SW(not ex-Nokia)

My point is, if your going to make NEW hardware.. why not start with a NEW linux distro and make it really open?

I would pay 700EU for a linux phone with HW keyboard But for 500MB RAM and locked into freemantel's aging repo's... edge/jolla is looking a lot better.

While I respect your preferences, that's all they are. To say that something without a keyboard is crap is completely unquantifiable without using an objective set of qualifications. Something can have attributes and characteristics or not. The same holds true for stating that Fremantle is better than Harmattan. The same also holds true for the form factor.

For each use case, an attribute/characteristic of the device either aids or detracts from the ability to execute the use case.

I personally find myself using the software keyboard on the N950 in portrait mode more than the hardware keyboard. The screen is bigger than N9 so soft keys are easier to hit. Also, it's a pain in the *** to flip it open and close all the damn time.

I'd love a portrait mode Fremantle that works great with no keyboard with a slimmer line. If the board that is built takes into consideration that I may want to shave the back off my N900 case and throw out the keyboard, that'd be great.

Keep in mind the N900 virtual keyboard isn't exactly very user friendly, when compared to most other virtual keyboards out there. (key alignment playing a big role here). Also, that would make it impossible for you to search for a certain name in the contacts application, etc. Many apps do not offer "alternatives" (such as a search button) for the lack of a hardware keyboard, since it makes sense to use the keyboard on the N900.

please note that one of the absolute axioms of this project is: we can do it because we already have a "protoboard" called GTA04 we know how to build, that is already very very close to N900 (NOT N9, NOT any cutting edge competing with galaxy 96). We're not aiming at bazillionth single board system ala rPi, we're not inventing a new phone and then find a matching OS for it or even create a new OS.

The GTA04 fits into N900 case and is already similar to N900 hw. We will slightly augment it and give it more ooomph where we can do. Then we're going to sell those devices to a community that's interested in extending the lifespan of their OS which happens to be maemo. We can't tell that community to pretty please now port ubuntu to our new leete device which happens to look like N900 but is something completely different inside, with latest cheap SoC from Samsung or whatever.
I thought I made all those points clear already in previous posts of this thread, but obviously they need to get stated again, more clearly.

Community task force are planning to port fremantle to Neo900 since what we have in maemo community is maemo fremantle OS and that's known to work kinda stable and power-economically on a platform like N900 or GTA04. And for fremantle all the urgently needed apps like dialer etc and all the middleware is existing and working, albeit much of it closed source. We're going to change that long term (buzzword "freemantle" - I don't like the name) - see CSSU - but for sure we're not doing anything like "rebasing fremantle on debian" or whatever weird other ideas might come up. Nobody is going to accomplish such task in any reasonable time frame, so the project would already be doomed from beginning. And no, you definitely can't port maemo harmattan to the Neo900 and expect a working usable pocket computer with pretty good phone capabilities.

If you want a rebased-on-whatever-OSwhatever, on bleeding edge hw platform available for 250EUR, go ahead and pick any arbitrary existing hardware available out there and port the OS of your dreams to it. You don't need neither GolDeliCo and GTA04 nor the N900 case for that, though you're of course free and highly welcome to buy a Neo900 and do to it whatever you like :-) And you get all the docs to empower you to do so. There's no such thing like "locked into freemantel's [fremantle(!)] aging repo's" on Neo900. If you want to use debian bleeding edge, just go ahead. there's a lot of alternative distros that get ported to Neo900 pretty quickly and some are already ported to N900.

cheers
jOERG