News:
19/07/2004 - The developement has reached a really nice stage....:-)
Other changes were made to the bayer to rgb routine; Tomas has ported the driver on 2.6 kernel;
the (famous) allocation bug has been fixed on 2.4 kernel (it is still in 2.6); I started to work at
the code for support all the icm532 resolution; a prototype of auto-brightness function has been implemented.
As soon as this new functions will be working a new tarball version will be released.
In the meantime the code is available on sourceforge's cvs.
04/07/2004 - Rev.0.3 is out! I finally found some time to work again on this project....:-)
Changes were made on the bayer to rgb conversion routine originally provided by Takafumi.
Other "improvement" were made on the start-stop frame detection (I simply abandoned the find tag
routines and moved to a fixed intervals approach).
Take a look at the latest screenshot with camorama at my Tux.
23/06/2004 - Thanks to Tomas Groth's message I discovered his great work on icm532 cam.
17/06/2004 - Rev.0.2 is out! Thanks to Takafumi Mizuno a new uncompressed data mode have been implemented.
Great improvement: images are quite recognizable! Take a look at the latest screenshot with camorama at my Tux.
The start-stop image and line are to be improved for avoiding all that flickering.
25/05/2004 - Looking for someone who may help me in hacking the windows driver in order to find the decompression
algorithm to use.
If you feel confident with such stuff, please email me and I'll be pleased to give you all the needed informations
20/05/2004 - Finally this site is up and all source is available for download!
29/04/2004 - The driver is now able to continuosly receive frames from the cam and to pass their raw data to applications
(see a screenshot from camorama).
The last step and challenge will be to find the proprietary decompression algorithm, without it data have no sense
- Removed part of all the unused (or previously used) variables in order to fix some kernel oops
due to unfreed memory. There are still some of this bugs, it's not my goal at the moment...
18/04/2004 - Moved all the declarations in a separate header file (icm532.h)
- Moved the icm532 strictly related functions in a separate file for better focusing on them (icm532_camera.c)
- Fixed the strange way I was using for resetting endpoint before submitting isoch urbs
jan. 2004 - All started here...
Taking a lot from examples (usb-skeleton.c) and real drivers (vicam.c) I started
to write icm532.c, a simple kernel module that tries to reproduce all the usb traffic seen with the windows driver
through usb-sniff