sicherheitsschwankung.de - some things I do for free

After having read about the board a bit and getting courious, at LinuxTag 2008 i could try the real hardware for the first time. They had a demo running which showed a full Enlightenment desktop and everything seemed quite fast. I would guess the performance is similar to a P3 1GHz. The demo was based on OpenEmbedded, but with a system that is as fast as a few year old desktop system, i would of course prefer to run Debian.

I've just booked my plane tickets for my first DebConf and if everything goes according to plan, i'll be there for the whole camp and conference. During the camp i'll work on getting E17 into unstable and on kvm/libvirt integration if there's still time left.

Some of the E17/EFL libraries are currently available in experimental. E-DBus and Efreet are next, then we can upload E17 itself.

Thanks to a patch by Roman Moravcik i was able to close OpenMoko bug 677 today. This problem made the SD slot unusable for anything except some simple tests. To verify that everything is working now, i installed the new Debian ARM EABI port. So now i can use Aptitude to install almost all Debian packages.

I'll publish a tarball of my chroot soon for anyone who wants to try Debian on the Neo. If you have problems with the speed of your SD card, check if the filesystem is mounted 'sync'.

Update: The tarball is now available.

Today i've recieved a mail from James Troup saying that my Debian account has just been created! This means my NM-Process took about 16 months (6 months in "Awaiting DAM Approval").

I'd like to thank Christoph Haas as my long term Sponsor and Advocate. He has been a tremendous help with all sorts of Debian and packaging releated questions. I also want to thank Christoph Berg for being my Application Manager and making that part of the NM-Process fun. :)

My first upload will be for BitBake to include a patch pull from upstream SVN.

While reading the OpenMoko commitlog I noticed this patch by Harald 'laforge' Welte. It updates the initialization routines for the GTA02 (the next Neo1973 hardware revision, Phase 1+ and 2) and creates a driver stub for a chip called 'SMedia Glamo3362'.

According to the manufacturer's site and datasheet, the Glamo3362 is the VGA variant of the 3370 which supports 2D/3D accelleration and MPEG4 en-/decoding (it also has a camera interface, which probably won't be used in the Neo1973).

Intel has just released a new utility called PowerTOP. It uses the new statistics made available by the 2.6.21 kernel and shows which processes (or devices) cause the processor to wake up.

Keith Packard said that he managed to increase the battery life of his Panasonic R4 laptop from 4 to almost 7 hours.

I'll try to reproduce his success with my X60s. If you want to do the same and use debian, it is already packaged uploaded.

The latest Chaosradio Express podcast (#42) is about the OpenMoko project. Tim Pritlove and OpenMoko core developer Harald 'laforge' Welte talk about the project and its challenges, the Neo1973 phone and open phones in general.

Harald said that the Phase 1+ hardware will probably include an upgrade to 1GB on-board flash.

We now have a game working on the Neo: SuperTux

Sound and music are not yet working and its impossible to control Tux (because we would need more buttons ;). Perhaps we can use a Bluetooth gamepad.

The OpenMoko websites are available:

• Main Page
• Wiki (see "Disassembling Neo1973" for some photos)
• SVN
• Some new lists have been created.
• Bugzilla

Happy hacking!

Sean Moss-Pultz has just announced that the Neo1973 hardware has been delayed by some necessary changes and bad luck. The software will be released anyway, so we'll be able to take a look at it soon.

I had hoped to take it with me to FOSDEM, so instead we'll work on OpenEZX some more. Perhaps we can show you something working. :)