Re: The new QWERTY device project
 

+1 for Sailfish being the first citizen. Handsome linux OS, It will also be good for Jolla to boost Sailfish, having a actual device that is on sale.

Android should be 2nd option for the masses, so Chen and Company deservely get rich and keep boosting handsome devices.

Third OS I think KDE plasma mobile should step in, Ubuntu is not in the race anymore, and I dont see any other distro advertising that they are going mobile. And this project perhaps could be a incentive to them to be faster on development, if their resources permit it.

Re: The choice of OS of the new QWERTY device project
 

@chenliangchen and others, there seem to be some misunderstandings in detail, ...

No, Nemo / Mer / SailfishOS is a GNU/Linux (i.e. an OS based on Linux (i.e. the kernel) and GNU utilities) and some community Distributions (SHR etc.) are, too. And Ubuntu mobile / Touch / for phones was GNU/Linux based as well.

But the development of all (nowadays) purely community-driven, mobile-oriented GNU/Linux-Distributions (e.g. SHR, Maemo5 / Fremantle, MeeGo 1.2 / Harmattan) has stagnated for years, and Nemo / Mer / SailfishOS are the successors of MeeGo / Maemo, as I already pointed out here / see Wikipedia.

No, most manufacturers of (ARM, MIPS etc. -based) SoC do care (but in greatly varying degree) about upstreaming their drivers for the basic functions of their SoCs, e.g. GPIOs, basic Audio, serial buses (RS232, I2C, SPI etc.).
The two crucial points are the video (encode, decode, 3D) drivers and drivers for special peripherals outside of the SoC (camera, touchscreen etc.): These device drivers are often not upstream (i.e. in Linux, the kernel) and sometimes proprietary.
That's why SailfishOS and others (e.g. Ubuntu Touch did) use libhybris together with "Android kernels" to be able to use these drivers, but this is technically not a must (i.e. a mainline kernel could be used, if it supports all hardware of a device and no Android runtime environment is needed).

Basically yes, but Nvidia's Tegras are designed for mobile gaming and tablets, thus their power consumption make them unsuitable for phones.

But you stated, that you already settled on a MSM8953 based design, which is an excellent choice IMO, WRT the Qualcomm Snapdragon 625 (MSM8953) SoC *and* the fact that you are building upon an existing, working design.
Are you really reconsidering this decision?

There already is one, called Freedreno (for Qualcomm's Adreno GPUs).
It has become basically usable recently, in contrast etnaviv (for Vivante's GPUs in various SoCs) and Lima (for ARM's Mali GPUs, but the Lima development has stagnated for a while, already).

But taking care about this and its integration, is the job of the makers of GNU/Linux-Distributions, e.g. Jolla, RedHat, the Debian project, Canonical, SuSE etc.
Do not underestimate the efforts it takes to create and maintain this.

Hence, please focus on creating a device with a well documented and open boot-loader (to make the device "hackable", so one can install whatever OS), while offering at most two choices of end-user-ready OSes preinstalled.

P.S.: While "regular" (i.e. non-mobile focused) GNU/Linux Distributions for ARM processors (Fedora / CentOS, Debian, OpenSUSE, Ubuntu etc.) on handheld devices are "cool" for nerds and hackers, they lack a telephony stack (i.e. no calls!) and applications suitable for touchscreens and for very small screens, making them a really bad fit for an end-user phone.

Re: The new QWERTY device project
 

Just to share another peek from the top. A lot design work is in progress so don't regard it as final version yet. Still very early.

Re: The new QWERTY device project
 

I would love native compatibility with full-fledged GNU Linux distributions, This would seriously be an amazing achievement and this would make the phone a real computer (especially with OTG and an USB-C hub). However, if this turns out to be a blocker because of too many technical difficulties related to hardware/drivers availability, I would still consider Sailfish OS + a good chroot (like EasyDebian on N900) to be a success and to be far superior than whatever can be considered competition to that phone.

Of course if it turns out that natively booting in a full-fledged Linux distribution is an achievable goal, then yay, go for it, but if it turns out being a lost cause and if this goal compromises the whole project itself, remember chroot will not provide the same advantages but would still be an appreciated and useful asset.

Re: The new QWERTY device project
 

+1 for Sailfish OS. I'm using it for two weeks and already forgot my Iphone 5s. Combined with a great machine and i'm sure it will have success

Re: The choice of OS of the new QWERTY device project
 

From my point of view, this is kind of a "if everything worked, I would prefer to run this-and-that OS" kind of comment. As to why: for me personally, a distro like Arch or Fedora is much more geared towards productivity, which is what I want to use such a device for, after all.

That being said, if it doesn't work well, I'd probably prefer to just use Sailfish.

Yes, I agree that this is the correct (and probably safest) option for Chen at this time.

Yup.

Chen, I think this is the best solution for you: offer Sailfish (for Jolla and GNU/Linux fans) and Android (for the average consumers). And allow the users to unlock the bootloader (for power users), so they can experiment with other OSes if they want.

Thanks for all your efforts!

Re: The new QWERTY device project
 

Atleast there's support for Adreno 506 GPU from freedreno: https://www.phoronix.com/scan.php?pa...o-A5xx-Bringup

So if we get the touchscreen, keyboard and wifi working that would gibe me an usable device :) Camera would be bonus.

Re: The choice of OS of the new QWERTY device project
 

Thanks a lot for this.

No I'm not considering to change SoC. What I asked is the reference for future development.

The ultimate goal is to have a GNU/Linux that is open to *any* development, like what Maemo nearly does 10 years ago. Sailfish OS is the most mature choice at the moment but our journey doesn't end here. This is why I want to know all the possibilities that can be achieved in the future.

Telephony support isn't a big issue IMHO. I agree most applications won't run properly on a small screen (Although a lot thinks 5.5 inch is way too big) but I think that's mainly because there is not enough proper handheld Linux machine at the moment. As the device choices grows more applications can be adapted.

(Maybe can also make a stylus for capacitive screen)

Re: The choice of OS of the new QWERTY device project
 

I did not even dare mentioning the stylus in this thread because it's not so useful with Sailfish, but if you have productivity applications running on the device one day (either with a full-featured Linux distribution or with a chroot allowing running desktop applications within Sailfish as EasyDebian did with Maemo), and you manage integrating a stylus within the body of the device on top of that like the N900 stylus, then I'm afraid I'll have to send you love letters every day. Are you. Really sure. That you want this to happen. I mean, are you.

(Not sure about capacitive styluses though, I have yet to find one that really provides the same satisfying experience as on a good resistive screen. I wrote a lengthy post about why I think stylus and stylus-friendly UI/devices are a game changer in the Neo900 thread.)

Re: The choice of OS of the new QWERTY device project
 

This comes another benefit of screen being bigger, we can then have possibility implementing a stylus slot without compromising too much on the structure. (And not make it too thick as N900)

A stylus like Galaxy Note series could be great.

Also remember N900 is 3.5 inch, on a 5.5 inch it may not need to be *that* accurate. :D

But anyway these are the possibilities for the future, will try to keep the topic on the current one.