T-Mobile with ASU on GTA01?

thomasg thomas at gstaedtner.net
Wed Jul 9 16:38:13 CEST 2008


 You surely are right in some points, you might even be right with the
broken modem (still trying to figure that out), but I strongly disagree with
you in the most other points (off topic, but I'll reply anyway).

On Wed, Jul 9, 2008 at 10:23 AM, Marcus Bauer <marcus.bauer at gmail.com>
>
> You may have one of the broken GSM modems. From my four phones two have
> a broken GSM - that's 50%.
>

My 1973 seems to be fine, I did calls with all images on the Freerunner,
too, and can't tell what are software-problems and what not.


> They constantly reconnect to the cell tower and inbetween the can't make
> phone calls and loose the GPRS connection but without notifying the
> pppd.
>

 Can't confirm this for 100% of the cases, I did some calls on the 1973 over
the last year, even ones with some minutes and it was fine. The main problem
was the miserable sound quality. At least that seems to be hardware-fixed in
GTA02.
Didn't test much GPRS, but had no loss of connection yet.

This is the same with the mature OM2007.2 images as well as with the
> professionally by Trolltech developed pure qtopia images, the ASU images
> the the new hyped FSO images.
>

Imho the OM2007.2 images are all but mature. Even ASU is more predictable
and reliable (except the Qtopia on X11 parts).
Qtopia is ok, Qtopia on X11 is a mess and still not really working most of
the time.

This is another hardware problem which is shared with the Neo 1973 and
> thus known since a year. The answer by Dr. Michael Lauer was "Guys, this
> is a Heisenbug. We pray that it does not occur too often in the field.".


I wonder that I haven't heard of this yet, but let's hope they are figure it
out and fix it.


>
> That is a very interesting engineering approach...
>

:(


>
> The bug in question is:
> http://docs.openmoko.org/trac/ticket/1024
>
>
> The big problem with Openmoko is this "not invented here" mentality. The
> OM2007.2 images were working well, GTK is a valid platform for mobile
> gadgets (see Nokia N700, N800, N810), you can add "bling" (see clutter)
> and there is a huge developer base. The qtopia port to X adds a second
> huge developer base.
>

Like said before, they are (at least for me) all but working well (not even
to mention the UI design with 2-color icons without ever knowing what they
might mean).
I don't see how GTK will bring the huge userbase. Nobody knew about
EFL/ASU/FSO from the release of neo1973 until december, but there barely was
any development.
In my opinion the reason is, that there was no real API, mostly not even
backends (except a messy GPSD and GSMD).
The FSO-stack is what's needed to get that API and to get the developers.
They don't want to have a 1-Toolkit-Strategy, they mostly want a nice stack
to work with and use for their frontend what ever they want. Btw. - the main
developer of the so called "Underground" realized that a year ago.
The userbase of Nokias Internet Tablet Series is huge, but it looks for me
as the developerbase isn't that big in relation, and that with a GTK-stack.
Even Nokia bought Trolltech, and I bet they'll drop GTK eventually (and this
is a huge sign, as GTK is LGPL and so much more interesting for commercial
and closed source apps than Qtopia with GPL and commercial licenses!). They
are paying the Trolltech folks now, and they are also paying EFL developers.
Looks like the CAN see beyond their own nose.
Qtopia doesn't add a huge developer base. Qtopia was used mainly in closed
environments, there is no open source developerbase. Note that I talk about
Qtopia, not about Qt. I like Qt as much I like GTK and I think both are
really necessary on the Openmoko platform (as much as I dislike much parts
of KDE and especially GNOME).

The biggest project with the NIH-syndrom is definitely GNOME. They do _all_
from scratch, they'd never ever use anything that's not from them. Hell, I'm
glad that there is FDO to coordinate between them and the (still imho) much
less arrogant KDE people.

The "bling"-argument with clutter can be just a joke, talking about a
platform without OpenGL capable drivers (and for the coming device even
without OpenGL capable hardware).


> But instead of going on and having a base for testing the hardware,
> there came this change to ASU and etk which probably 0.1% of Linux
> developers use. And despite what Lauer & Co try to make us believe, this
> alienates GTK and qt developers. Just look on the planets of KDE and
> GNOME - nearly no mentionings. The developer mailing list: a big void.
>

I don't see your point here (exepct that you are GTK- and maybe
GNOME-developer and don't like other platforms).
Etk behaves under Illume/Enlightenment just like GTK and Qt do. There's no
difference at all.
Nobody is forced to use Etk and nobody ever will. The difference is, that
you can use what you want without to be forced to use something specifically
(like Qtopia without X11). Im also glad, that it looks like there is no need
for Evolution and GConf in future, I personally don't like them and don't
want to be forced to use them.

Again on the developers: they had nothing to work with. FSO with it's
dbus-api will make this much easier and everyone can use it with the other
libs he likes. And again: it was much more calm when GTK still was the main
and only platform.
I don't see any "alieniation" here. GTK, Qt, Etk, WxWidgets, GNUStep, ... -
all they can be friends here. None of them (this goes for Etk and Ewl, too!)
is used in the base-stack or the main gui (this goes for ASU and FSO the
same way).

I also don't fully agree with every decision openmoko does here, but the
development goes slowly in the direction I wished a year ago.


> But the real problem here is that basically due to this reinventing the
> wheel with ASU nobody inside Openmoko has ever really used the phones
> thus plenty of things which could have come up simply got lost. If Sean,
> Wolfgang and Steve would have started to exclusivly eat their own
> dogfood, i.e. using the Neo as their daily phone, things like
> oszillating GSM modems, non working GPS, SIM cards, deep discharge
> batteries, noisy headsets would have been since long ironed out.
>

Reinventing what wheel? I'm pretty sure that the invention of the wheel like
we known it didn't count as the reinvention of the cube (this is what the
old GTK-stack is: a "wheel" with corners).
In the second part here you mix that with hardware issues, can't really say
anything about that. But I agree: the Neo is no all day phone (yet?) and the
bugs are a mess.
On the other hand switching from GTK in the new direction would have been
necessary a long time ago. Not because of the graphics, bling, preferred
toolkit, ..., but because of the backends and middleware.


>
> Before now all the fanbois jump onto me and accuse me of trolling:
> in order to come to some lifestyle competitor of Apple the important
> thing is that the basics work and that they work reliably. Accepting
> brokeness as part of freedom is doing a disservice to the free software
> world.


Nothing that was not mentioned above here.

And it is even more unacceptable as there were 5000 people buying a Neo
> 1973 more or less for nothing. They all would have been more than happy
> to participate in advancing the Neo.


Same here.

The point being: the Neo *is* a fantastic concept. Bring it there. Stop
> ASU, concentrate on the basics, get the gtk and qt communities in.
>

Agree, but let's stay at GTK and Qt communites, not the parts of them that
*hate* all other toolkits (and this goes for big parts of the KDE and GNOME
communites, they mostly copy each other to not have to use apps with the
other toolkit!).
It's about choice, not about telling others what to do.

This goes to the executives of OM Inc.:
"Stop ASU"? I don't know. It seems that there is someone who had the idea
and wants it implemented (and I guess this most be one of the core members
or executives), so please, let not only the developers who do it communicate
openly, do it yourself!
The persons who does the decisions in the software area seem to be very
quiet on the list - involve the community!
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.openmoko.org/pipermail/community/attachments/20080709/d8ee386b/attachment.htm 


More information about the community mailing list