Mer as a based OS for Webian?

[adi cahya]

Hi all..

i a little quite in here lately right?

Well..just playing around at meego.com. I found that the community is going many direction. Some of them decided to create "Mer".

This is the excerpt from the blogpost

"To get one thing out in the open: this is just the core OS, a Linux distribution. There is no UI, and hardware adaptations are seperate from that core OS. It's an extremely slim Linux vehicle for making products out of. What you put on top is entirely your business - it's just a tool."

The UI can be Meego handset UX, KDE brand new Plasma Active One..or maybe Webian? :)

the full post can be read here

http://blog.rburchell.com/2011/10/meego-reconstructed-plan-of-action-and.html

Adicahya 

[Ben Francis]

Hi Adicahya,

I was interested to hear about both Tizen and Mer, and I think Mer has also been suggested as a base for Mozilla's B2G project (they're currently using Android as a base). It will be interesting to see what the Mer project comes out with (I think a useable OS is probably quite a way off yet).

I'm actually currently thinking that while B2G focusses on the mobile space, perhaps Webian should focus on nettop and netbook form factors, perhaps converging towards tablets and mobile devices in the future. In this case a more standard OS like Debian or Ubuntu might be a more suitable base to begin with?

Ben

--
Ben Francis
http://tola.me.uk

[adi cahya]

Well,its up to you Ben.

Anyway,in my studio,we use a system called AHP to make a selection from different choices.
We first pick the requirements,and ranked it. Then we compare each alternative (in our case,the based distribution,ubuntu or else) in each or requirement.
At the end,we pick the alternative with the highest score.

So..what are the requirement for the webian based? Portability,active community,stability,scalability,wayland support or else?

Maybe we should discuss these first.?
If u interested with the AHP I can send u the template.

Asicahya

[Jono]

I had never heard of Mer up until now - it looks promising, but as you
say something usable (with good driver compatibility) does seem to be
a way off ...

> In this case a
> more standard OS like Debian or Ubuntu might be a more suitable base to
> begin with?

I think that's a good idea to start with, although it depends how long
it is before Webian is usable as an OS. My feeling is that it should
have a custom/minimal window manager based on either X11 or Wayland
(do Mozilla's frameworks support Wayland yet?) with the browser
running on top (or have the WM integrated into the browser, but I'm
not sure whether that would be much of an advantage). This would mean
it would work with any distro as a base - you could start out by
making it a Debian package, and then move to Mer as a base later on if
it looks promising.

The main advantage I can see of using Debian/Ubuntu is driver
compatibility - Webian will need as many wireless drivers (in
particular) as possible because everyone hates building from source
>_<
IMO Mer will only be a good base if it supports the same amount/more
hardware than Debian.

[sam smith]

I agree with Ben, I've been doing some research and in depth pondering
over time and I think linux(Debian ARM dev to be exact) is the way to
go. Creating one main footprint capable of running on ARM devices
would be far easier and efficient development, scaling and run-time
wise. There are tons of ARM mobile devices and it would be simple to
port it through QEMU for x86 devices and would much more efficient
than vice-cersa. Debian is one of the best option. Ubuntu even if it
were more stable, didn't have all of the split issues and worked
better with the linux community would still be a poor choice in
comparison with Debian. The only issue(as which has been discussed
before) is the Xorg and I think if some type of touch interface
friendly universal UI was create it would be the solution to a lot of
problems. Things such as android and other alternatives are become
evident as the half-baked "wait while we fix this mess we create"
corporate solution to customers they only value as profits as opposed
to users in a community who knows what they want.
This also ensure responsiveness and graphic capability. I have a
Tegra2 tablet and would love to assist in some type of solution in
this category, though last I check nvidia was hanging tight to the
source for no reason.
Debian ARM branch would be the best option with a custom UI or
modified Xorg would be the best route as well as have the most
resources. The most work would be the kernel and the UI and the former
I imagine would have plenty of material/experience/advice in android
community custom kernels. I don't know how much of the new gnome3/
shell but its seems to fit really well as well as there are tons of
extension etc. that can also be benefited from.
Debian Testing (ARM)
Gnome Shell
and a custom/hybrid android kernel, or perhaps something even better
if Nvidia gets on track.
Obvioiusly there will be things to tweak such as boot, memory
management, etc. but keeping it linux and the community
Just my opinion, feel free to tear it apart and call me an idiot, I'm
open to criticism and education.
Sorry about the long unwieldy rambling post.

> >http://blog.rburchell.com/2011/10/meego-reconstructed-plan-of-action-...
>
> > Adicahya
>
> --
> Ben Francishttp://tola.me.uk

[Ben Francis]

Sam,

I like the idea of targetting ARM, I'd really like to prototype on power efficient ARM-based nettop/netbook devices. I know that the whole mobile world runs on ARM, but do you know of any good ARM hardware for nettops/netbooks that we could target?

I think x86 is probably the more popular choice on these form factors at the moment.

Ben

[carli]

Use emulators. As a "real" ARM device I have the OpenPandora (http://
openpandora.org) which is not as comfortable to develop on as a
desktop computer.

[Jono]

I also like the idea of ARM, and I think it'll only gain market share
in desktops and notebooks in the years ahead. A quick Google found the
"eCAFE Slim HD" and Ex HD netbooks (sorry, can't remember the exact
page), which may be great or may be horrible - they have rather low
specs (compared to branded tablets) but this may not be an issue, I
have no firsthand experience. Maybe worth a look though if you're
interested?
But as carli says, emulators are probably the cheapest and easiest way
to go about it.

[sam smith]

@Ben: I believe that x86 will only be for servers/high performace
desktops and the ARM for low power, mobile, lightweight/standard
desktops such as Webian will become the future of computing. I feel
ARM targets the same audience and interfaces as Webian and by using
this down stream it would be simple to append Webian to any linux
operating system through the kernel with option to boot it or the
initial OS depending on the users need. This makes it easy for it to
run on Android, Linux computers and even windows systems with grub
installed. You could even then run a linux container on x86 to emulate
at near real speeds, that with the additional clock speed from the x86
should make it even faster than on ARM. Its like the best of both
worlds although you do sacrifice performance on older PI&II processors
but I doubt your targeting such a old market. Webian really doesn't
require a multicore, sandybride, or bulldozer and would add complexity
with compatibility. You should choose the architect that fits your
needs, if you use Debian theres already http://www.debian.org/ports/
for you to build on top of and get documentation. Not to mention
theres millions of arm devices for it to run on and building smaller
is key to web infastructure. However I have never built anything on
arm just android/java and kernels.
@carli: Thats what the QEMU/kvm is for, I can/open virtual instances
in a single command. Of course there virtualbox, vmware... etc.

[adi cahya]

i think arm architecture will play more important role on the near future.
so,if webian want to select its core,support for arm should be considered.

today,arm based laptop.is still rare, but i think when windows 8 come out.we will have more choices.

if webian release timeframe is close to win 8, then we really should considered ARM.

Just curious..can we apply some kind of developer version of ARM board from Nvidia maybe, or Qualcomm?

Right now, maybe we can use Asus eee transformer as a development platform. The form factor can be use for testing the usability of the UI design.

If you think we should go with x86 first, thats ok with me. just prepared the system to able ported easily to ARM in the near future.

Adicahya.blogspot.com

[sam smith]

I have the Nvidia Tegra 2 dual core, theres a decent community and
development behind it, I even overclocked to 1.5 with it barely
getting warm. I believe there's quite a bit information as well as a
small form factor desktop(stand. router size) that ran ubuntu and
android. I'm sure its the same for many other arm processors. Theres
tons of hardware too, and I've seen a number of labtops(granted mostly
from the Eastern market). The only reason it seems labtops are rare
here is that they are in the process of mixing color e-ink touchscreen
tech w/ and combining the tablet with the labtop to completely screw
up a genius mashup of the sake of "business".
Buts its like you say Webian really has little room in the x86 world,
if your running web apps most of its probably in the cloud and is not
going to use their intel iX cpu. Webian seems very touchscreen
friendly as well to which most devices are predominately arm.

[Ben Francis]

On Fri, Nov 4, 2011 at 10:51 AM, sam smith <y2gs...@gmail.com> wrote:

I have the Nvidia Tegra 2 dual core

What is it plugged into? One of these development kits? http://developer.nvidia.com/tegra-development-kits

[sam smith]

What I have is a Viewsonic Gtab tablet, the screen sucks but its a low
price android tablet with a overclockable dual core tegra2 so it
suffices for indoor uses for now.

On Nov 4, 8:27 am, Ben Francis <b...@tola.me.uk> wrote: