I'm a recent n900 convert and at this point I don't know if I can ever go back to Android/iOS. I'm eagerly awaiting the release of the neo900 and I honestly think this device is the closest we'll come to a truly open phone that you can use everyday.
I encourage anyone interested to consider supporting Joerg, Werner, Sebastian and everyone else working on this project! [1] 100 euro now and at the very least you can have a really cool device to hack around with.
if you donate 100 euros do you get anything or is it just supporting the project? Couldn’t quite figure out if by donating you got like an early revision type thing?
If you donate at least 100 euros, you can be sure that there will be a device produced for you (it's built to order, not to shelf). Also, this amount you pay now will already count to the final price, so if you donate 100 EUR now, you'll pay 100 EUR less when finally ordering the ready device.
Nice to see someone is attempting this. Physical keyboard and resistive touchscreen seem like pretty odd decisions though. I guess this is going to be as enthusiast-only as the previous Neo phones which is a shame.
s/odd/sane/:
This oddity you've identified is precisely the reason people like me are interested or even invested in the project.
Must everyone worship at the altar of smudgy screens
and ineffective text input? The N900 is the last phone
that was actually usable. And by usable I'm not only referring to the stylus/keyboard input duality, but also to the the fact that it's as trivial extendable as any reasonably open Linux system.
In fact, I'd wager that the two are very closely correlated:
the need to tinker and the desire to do so with sane input devices.
I hate capacitive screens. They're great for tablets, but overly limiting on phones with small screens - especially "hacker-oriented" ones.
I used to code using on-screen keyboard with fingernails on my old Neo Freerunner. It maybe wasn't the greatest programming experience ever, but it worked for me and I had some great debugging sessions while in tram :) I can't imagine doing that if Freerunner had capacitive screen - even if there wouldn't be a bezel around the screen anymore.
People seem to hate resistive screens based on their experiences with poor ones (which is understandable, as most of them are pretty poor). However, a good one is a real pleasure to use, and I find the one used in N900 a good one.
The N900 supports both touchscreen and physical keyboards, so you get the best of both and I'd wager the Neo900 continues this.
Personally I've always found the physical keyboard to be far more productive and have wondered why smartphone manufacturers didn't keep on making them.
Within a few years I think we'll see an open source phone similar to Bunnies novena laptop project. Imagine a crowd-sourced phone becoming the #3 contender in the smart phone wars, Android, Apple, and FOSS.
I never got the purpose of projects like this, at least they admit that there is no current way of guaranteeing a safe device.
Even with "open" firmware virtually any commercial hardware can be either backdoored or exploited. Modern phones (as well as any other computing device) have so many software components that i suspect intelligence agencies do not even have to exploit the main user OS to get data of them.
Modem and BBP OS, CPU microcode, bootloaders, firmware for other hardware components ranging from the camera to radio's not covered by the BBP, USB/OTG controllers and more.
I have serious doubts that "sandboxing" hardware components will actually work, especially in any useable way.
For example modems behave badly all the time, both due to internal glitching and due to environmental reasons. Heck dump your BBP logs from your mobile device and see how much shit it pulls off regardless of what the OS tells it to do because it's manufacturer thinks they know better (and they do most of the time), so every time there is some BS signal on the same frequency or the modem decides to wake up or hop channels on it's own the device will reboot? GL making a single call with that then.
Not to mention that i really don't see how they are going to do monitoring effectively, from the schematics they are doing power management monitoring which is nice but most BBP packages can receive even in "sleep" mode. The isolation they are suggesting might also not really work, sure putting in "breakers" (which i assume are either some low voltage relays switches or resettable fuses) and cut the power off completely to the BBP might work but cutting off the antenna from the BBP? doubtful RF signals should be able to hop most in circuit breakers, and many BBP modules these days come with integrated antennas for at least some of the low power/low gain RF stuff like BTLP and NFC.
To me this seems like another one of these "experiments" that will only result in overpriced and outdated hardware with poor user experience. I would love to see an N900/950 modernization just for the ability to run actual linux on the phone to serve as a toolkit. But if anyone thinks that any device they can hack together will give them any level of privacy over organizations like the NSA is living in la-la land...
>I would love to see an N900/950 modernization just for the ability to run actual linux on the phone to serve as a toolkit.
That's what Neo900 is. All privacy-wise "features" are just a side effects of designing it properly (as in: treating blackboxes as blackboxes instead of explicitely giving them full control over user OS, controlling what data the baseband has access to etc.). There are some additional measures like power switches, monitors, breakers - but those are pretty cheap to put in.
Nothing will reboot "every time there is some BS signal on the same frequency or the modem decides to wake up or hop channels on it's own". The point of Neo900's monitoring is to let the user access to information that modem is doing something. Of course if you really want to, you can then setup your software to automatically reboot (or power cycle the modem alone) - but outside of very special use cases, that would be pretty pointless. If you don't care, you simply don't use that data.
>But if anyone thinks that any device they can hack together will give them any level of privacy over organizations like the NSA is living in la-la land...
Neo900 won't magically protect you from anything - no device will. Instead, it gives you tools which, when used properly, can allow you to fight off some of the mass surveillance techniques. Like encryption, which is pretty much rendered moot in most of modern phones where the modem has direct RAM access.
Keep in mind however that when you are an explicit target of some three letter agencies, they can probably employ better and more effective methods against you than sniffing on your mobile phone.
tl;dr: Neo900's privacy solutions are rather targeted towards Stallman, not Snowden or some intelligence agent from foreign country.
Again i went over most of the schematics and the specifications and nothing actually in this project seems to be worth while.
The privacy features are kinda moot and not effective, and the hardware is a joke.
Low end SOC(TI DM3730 which is 4-5 years old now) with poor media support and very low performance and only 1GB RAM.
Limited RAM and extremely limited ROM/Onboard FLASH which will limit the OS in the long run or force you to use the SSD card which will lead to probably the most common leakage.
Oh and the screen? 3.5" TFT, 800x480 resistive, seriously where did they even find a resistive touch screen these days at the bargain bin on Alibaba?
And they expect people to pay 850 EURO's for that?
So no this is not a modernization of the N900 it's pretty much the N900 with hardware that is only slightly newer than what was available when the N900 came to the market.
There is seriously no reason for this phone to be so poorly designed especially with their mentality of sandbox rather than trust.
We had quite a few phone projects that were either designed by the community or by small companies and all of them managed to use top of the line hardware.
You have Snapdragon 800/801 phones under 300$ these days, why use a SOC that will not even be able to play a movie these days?
So sadly this isn't the driod you should be looking for, still waiting for a Tizen/Meego phone worth while Samsung has been shipping some in India, and Yatta has a very nicely designed phone out in Europe and Russia but none of them are toolkit phones sadly.
There is still a lively community around the n900, as certain folks love its features. Though it has some drawbacks like only 256MiB RAM, no BT 4.0, no LTE, propper USB OTG, no gyro, no compass.
The point of the neo900 project is to perserve the features of the n900 and gently deminish the drawbacks. It is not about having a octacore and 16GiB of RAM which people using the n900 don't even care about.
For me the most important features of the n[eo]900 are:
modem properly seperated from CPU/RAM (not the case for basically any smartphone of today),
hardware keyboard,
propper linux environment
I have an N950(bought of ebay) and i care about meego/maemo/tizen/fork of the week...
And in general that is nonsense, you've just mentioned the drawbacks of the N900 so why release a device which has the same drawbacks?
The SOC was lowend when it came out, and currently you could not find a smartphone with a SOC which even comes near it's performance even if you look at the pre-paid bargain bin at walmart.
We are literally talking about a SOC which will have issues playing a 1080p movie.
Why only 1GB RAM(other than the fact that the SOC they've chosen seems to not support more than that)? RAM costs absolutely nothing these days, the modems they've selected is also quite questionable considering it's made by a subsidiary of Gemalto especially for a "privacy" oriented device.
And 512MB Flash? this just puts unnecessary limitation on future developments of the Phone, which also means that it will probably not be well supported by the latest Meego forks or any major Linux distribution.
BTW according to their latest schematics i don't see any decent level of separation between the BPP and the SOC not on a hardware level at least, it seems that for the most part they still rely on the software to bring the modem back in line which well is iffy to begin with.
Just based on the hardware they put in the phone there is no reason for it to cost 100 EURO's not to mention 800-900. They've could went with decent hardware and still probably charge as half as much.
P.S.
Considering they use a SOC with a PowerVR GPU which uses a closed source linux driver and some other closed source drivers from what it seems i think they should change their tag line to "a mostly free software phone" :)
That driver issue alone should make the entirety of the privacy angle moot, it's completely closed and unaudited so for all you know it mirrors your screen so the folks at the NSA can see the porn you watch...
* "only 256MiB RAM" - Neo900 has 4 times more.
* "no BT 4.0" - WG7351-00 has BT 4.0 support.
* "no LTE" - you can choose PLS8 modem with LTE.
* "no proper USB OTG" - Neo900 has it, plus even internal USB pins for homemade mods.
* "no gyro" - Neo900 has one.
* "no compass" - Neo900 has one.
512MB flash comes together with RAM package, and it's actually nice to have just for better Maemo/N900 compatibility. There's no reason why you couldn't boot OS from eMMC, which will for sure be at least 32GB. Actually, if you plan to tinker a lot with the phone, I'd recommend having some sort of rescue OS in flash, and your own OS of choice on the first partition of eMMC.
1080p movie may be a bit tough to play on this SoC, right. 720p however should be fine, and I hardly see the point of going 1080p with 800x480 screen.
BPP and SoC are connected only via USB, audio codec and some GPIOs. It's pretty much as separated as it can be - maybe except using serial line instead of USB.
"They've could went with decent hardware and still probably charge as half as much." Of course - if we could produce number of units with a bit more zeros at its end, it would be very cheap. You haven't discovered anything special here - it's just standard economy of scale. With our current scale, any "modern" design would either cost ten times more, or wouldn't be possible to build at all due to lack of available components to buy in such low amounts.
That would be nice and all if the SOC could allow decent internet browsing experience and MM support.
So sorry i don't see how putting out a device with hardware which only slightly better than the original N900.
Yeah you'll have BT 4.0, decent WIFI and LTE but then again you won't be able to browse YouTube on that SOC properly.
I'm sorry but i really don't see the usefulness of a device with a single core 1Ghz CPU and 1GB ram these days and it's a damn shame, the whole idea behind Meamo/Meego is the freedom to do pretty much everything why put hardware in it that will not allow you to do anything?
Heck with that screen even office and text editing will be a bitch :(
Although I would prefer having a phone with an open-source GPU driver, the situation is not too bad with the Neo900: apparently you can run without the non-free blobs as long as you don't need 3D acceleration: http://neo900.org/faq#gfx
Please don't compare the OMAP5 to the dm3730.
While the OMAP 5 is not the latest and greatest it's still a modern SOC.
We are still talking about a SOC with a dual core CPU, a modern GPU and LPDDR2/DDR3 support.
The dm3730 has a weak CPU, supports slow and outdated RAM, and a GPU which will not be able to support modern multimedia.
Also Open Pandora is meant to be a dedicated platform for emulating (very) old games. Not a general purpose modern mobile computing platform.
For it's purposes the OMAP5 SOC they've selected is probably an overkill and most of the flak they got for it was from people wanting to use their "console" as a general purpose computing device...
As an option, no. It's too big change, it would basically be a different product. Also, being able to base the design on already working and tested GTA04 was a major point when we evaluated feasibility of the project at its start.
As an upgrade option later - possibly. After all, OpenPandora's Pyra is like our "sister project", sharing some of the developers. Making an OMAP5 upgrade board based on experience and knowledge acquired while working on Pyra shouldn't be that hard.
Why for? It won't change anything. On OnePlus the modem is tightly integrated with the rest of the system, and on Nexus 4 you cannot even boot Android to any usable state without proprietary blobs, so forget about any non-Android based OS (remember that Ubuntu for mobiles is just a chroot-like layer on top of Android base system).
I encourage anyone interested to consider supporting Joerg, Werner, Sebastian and everyone else working on this project! [1] 100 euro now and at the very least you can have a really cool device to hack around with.
[1] http://neo900.org/donate