Request for help: Would like community applications to show anddiscuss at LinuxWorld
marcus.bauer at gmail.com
Tue Jul 29 14:29:52 CEST 2008
thanks for taking the time for writing your answer.
On Tue, 2008-07-29 at 18:38 +0800, John Lee wrote:
> Part of my current work requires me to use fso daily. It seems
> strange that what I know seems to be different from what you know.
> * fso does not force you to ASU or closely connected in any way.
> could you please elaborate?
> fso: an open specification dbus interface (freesmartphone.org)
> + a reference design (frameworkd, check git.freesmartphone.org)
> fso-image: fso + a reference python UI based on EFL.
> asu: a enlightenment WM for mobile phone (illume)
> + qtopia phone stack (not based on fso)
> + installer (EFL)
> + diversity (gps app based on EFL)
> + exposure (config app based on EFL)
> the only similarity i can tell is EFL in fso-image. but the fso
> itself does NOT force you to use it, just the implemented reference UI
> used it.
As you note further down, OM is going to stick with FSO. Thus unless OM
is developing ASU just for fun, the assumption that it is being ported
to FSO seems more than vaild. Please correct me if I'm wrong there.
And as you note further down, phonekit needs to be ported, otherwise the
dialer and the sms-messages apps wont work any longer. This is not a
task one can do in an afternoon. Thus on the long run FSO effectively
forces to use ASU.
> it's easy to do another reference UI with GTK.
If Openmoko has taught one thing then the following: "easy" is nothing.
Otherwise people would buy Neo's instead of iPhones now.
> exactly what are tied together here?
gsmd and FSO dbus. The gsmd is the core part of a phone and if that is
incompatible to the current OM2007.2 one then dialer and messages stop
> for example, you can just run the ogpsd subsystem in frameworkd then
> use phonekit + gsmd to handle gsm if you want.
Which then will break ASU applications. And this is not how Linux works.
I can run Konqueror on GNOME or gimp in KDE or xfce or enlightenment.
> on the other way
> around, the frameworkd is just a reference design, anyone can take
> libgsmd + gsmd to make the same interface on dbus.
Again: "anyone can take" is not true. Anyone can take a couple of
transistors and make an iPhone - not.
> could you explain why it's a WTF idea to have a PIM API on dbus?
1) eds has already been ported to dbus - so FSO is reinventing the wheel
2) the whole world uses libraries at application level because it
provides a nice abstraction layer (and so does EDS-dbus). the difference
between a bus and a library is similar to a water bottle and a water
pipeline in the end the both transport water but they serve different
> there are technical reasons behind the re-implementation of gsm daemon
> but I'm not the one to answer it.
the gsmd works well. there is no technical reason.
> I think the reason why you are
> unhappy is that OM moved away from OM2007.2.
I'm living next to Sophia Antipolis which is the french 'silicon valley'
with 1300 companies and 30,000 employees. The common opinion here is
that OM shows erratic and unpredictable behaviour which makes it
unsuitable for consideration as development platform.
That's simply a pity. Unless OM wants to do everything by themselves,
they need to care for external developers too in order to set up a
working eco system.
> since OM will stick with fso in the foreseeable future, I think port
> OM2007.2 app suites to fso is a logical move.
If at all I place my bet on GMAE and would not recommend using FSO but
sticking with OM2007.2 which will give a much better exit path towards
Limo, moblin etc.
> ogpsd is there based on
> gypsy, and it should be just another backend of tangogps.
ogpsd should just offer the NMEA data on port 2947, thus keeping it
nicely network transparent. I'm not going to remove this functionality
from tangogps. Moreover the '800 pound gorilla' OM is developing its own
gps software and I'm not spending my energy competing with it.
OM2007.2 is there, it works and I recommend everybody to develop for it.
More information about the community