GSoC Interest

Mark Schneider mdschnei at
Thu Mar 27 04:51:51 CET 2008

On Tue, Mar 18, 2008 at 12:44 PM, Mark Schneider <mdschnei at> wrote:
> Dear Openmoko Community,
>  Since most of my experience is in hardware and low level software
>  (device drivers and kernel hacking), I think my skills would be best
>  used there, however, I would not be apposed to working in higher level
>  middleware.
>  Initially, I had wanted to write an open source device driver for the
>  GPS device in the Neo1973 that would provide a standard NMEA output
>  which gpsd could interpret.  However, I see that the Freerunner will
>  be getting a new GPS device, so this may no longer be necessary.
>  Other ideas that I saw on the GSoC wiki page that I thought might be
>  of interest:
>  Ad hoc communication via Bluetooth/WLAN
>  Cooperative Differential GPS
>  Accelerometer Gestures
>  My willingness to work on the project is not conditional on whether my
>  application gets accepted.  I would like to work regardless of Google
>  supporting me.
>  If there are any other projects that you think would be good, please
>  let me know.  I would like to discuss this more before I submit my
>  application.  Email works well, or you can occasionally find me on
>  #openmoko under the handle 'queueRAM'.
>  Regards,
>  Mark Schneider

Dear Openmoko Community,

Thank you to those who have responded to my questions.  After some
more thought, I would like to propose another idea for a project.  I
have seen that the Neo1973 takes a while to boot (~1.5-2 minutes).  I
would like to see if there are opportunities to speed up the boot
process.  I have noticed that there has been some previous work done
by Alessandro to profiling the boot process with bootcharts [1].  My
idea is to start with the kernel, to see where in the kernel there
might be room for improvement and then continue into the boot process
by using Alessandro's bootcharts work as a reference and coming up
with other ways to measure the processes that consume the most time
and try to work with them to improve their speed.  Before submitting
this in a GSoC application, i wanted to throw this idea out there in
case anyone had any thoughts on the matter and to make sure this work
hadn't already been done.



More information about the community mailing list