[maemo-developers] Nokia and MS

[From: Angel Perles]

Hi,

I think the point of view of Niels is the best way, but with some 
considerations.

I would like:

- an open platform
- based on (de facto) standards
- the posibility of earning money with it, so no a geek/exclusive/niche 
platform

In this moment Meego is out of my consideratons. We have been a kind of 
beta testers with the N900, so I don't want to repeat.

My previous smarthphone was a Windows Mobile 2003 HTC Blueangel.
I hated the lack of support of the supplier and Microsoft, and I was 
forced to benefit of the hacking efforts at xda-developers to get it 
phone working properly, and helped others to do so.

But doing the work for Microsoft and being paid with threats was not 
pleasant.

So I decided to start playing with Linux. I supposse that most of us 
remember, Angstrom, handhelds.org, Qtopia (yes, Qt), Open Moko, etc. It 
was great, but always a marginal platform for geeks.

I think that a good path is to mix the goods of Meego and Maemo. That is:

- development tools: support both Qt and GTK+, enhance Java support 
(including Dalvik, if posible)

- replace closed parts of Maemo with open pieces. Don't solve bugs of 
closed parts to force hacking if necessari (I'm volunteer for that)

- evolve Maemo to a kind of Meego. Meego is a good design, IMO. But I 
insist, don't try to get Meego working, in this moment it is better to 
get Maemo working well, and not to reinvent the wheel.

- make Maemo feasible on other platforms

- create a kind of economical ecosytem for this "oh!Maemo", so people 
can use it for their phones (paying?), embedded systems, buy gadgets, 
Apps, etc. (opps, lawsuits?).


Please, don't make the same mistakes of OpenMoko, Angstrom, and others.

I will maintain my N900 until it gets broken. It is not perfect 
hardware, it is not perfect sofware, but it is the best sum of both.

Àngel

>
> Isn't what you say here effectively the choice between two dead horses?
> even if Intel wants to make Atom/Moorestown based mobiles a reality I
> have my strong doubts in them. And if I am right they will never produce
> a real world Intel based handset which would then, without Nokia, make
> MeeGo's mobile UX dispensable - and under cost pressure it will be
> canceled, just the same way as they just canceled the netbook UX (with
> braindead reasoning but that's another story).
>
> I would then rather choose the horse/platform with most and best
> experience (Maemo5) instead of waiting for the other dead horse to grow
> beyond infantility.
>
> I think we now have the incredible chance of having a pretty mature
> platform (Maemo5) at hand in the open which we, the community, can
> further develop and extend without the need for any manufacturer
> support. We have devices and we have the platform. What else does it take?
>
> Concerning MeeGo I am extremely sceptical about its further development
> and almost as sceptical for Qt's future too. What will happen to Qt if
> Nokia reduces its effort into it? Nobody knows. But with GTK+ it is
> quite the contrary. GTK+ is a community effort from the early beginning
> and has an active community further developing it. Slower than Qt,
> admittedly, but it is working out. What happens when large companies
> dump gigabytes of sourcecode into the open has been proven a lot of
> times already - in 90% of cases the source starts to bit-rot.
>
> So I think we should free as much of Maemo5 as possible(*), put it into
> the open and ask Nokia for a guarantee of keeping the internet platforms
> (namely *.maemo.org) up and running or at least give enough time to
> mirror them to some other hosting service.
> And then make Maemo5 *the* true open source mobile platform.
> There are other devices evolving which could need it - google for GTA04
> for example. And once it has been ported to another device (for the
> timebeing I am not aware of any Maemo port to non-Nokia hardware) other
> manufacturers will surely jump - not everyone wants Android but it is
> currently the most portable and featureful platform. We could change that.
>
>
> (*) I am unsure how much of Maemo5 Fremantle is still closed. As far as
> I know it still contains quite some closed source components. A list of
> closed components would be good to have BTW...
>

-- 

*************************************************************
Angel F. Perles Ivars

Departament d'Informàtica de Sistemes i Computadors - DISCA
Universitat Politècnica de València - E.U.Informàtica
Cami de Vera s/n. 46022-Valencia
Edifici 1G Despatx 2S-13
e-mail: aperles at disca.upv.es
Telf.+34 963877007 Ext. 75775 Fax.+34 963877579
http://www.disca.upv.es/aperles