[sane-devel] reverse engineering advice.

[gerard klaver]

On Mon, 2005-01-24 at 13:59 -0500, Mbosowo I Sampson wrote:
> I'm in the process of sifting through USBSnoop generated logs for the HP 
> 3970, trying to decipher what's going on. I've hit a little bit of 
> a wall. Reading the code for the hp5400 driver along with the USBSnoop 
> logs posted at http://home.kabelfoon.nl/~bertrik/plugin_hp5470c.log
> has really helped a lot. I would like to start writing my driver but I 
> need some advice.
> 
> This is the log generated from my scanner after its plugged in:
> http://www.ic.sunysb.edu/Stu/msampson/usbsnoop_pluggedin2.log
> 
> The first 4 URB transfers are easy, querying the device, choosing a 
> configuration, etc... But after that it gets a little murky. I'm 
> looking at the packets going down, and coming up, and I just have no way 
> to telling what's doing what. I wanted to use usb-robot to send commands 
> one at a time, but the program segfaults on my machine (debian sarge 
> 2.6.10)
> 
> I have the USB1.1 spec handy, so I understand the structure of the 
> packets,  but how do you discern what the packets are doing? For example, 
> Which of  those packets turns on the lamp? Which ones controls the stepper 
> motors? Should  I just start from the top of the log and send the commands 
> myself to see what happens? I'm really itching to get this thing working.
> 
> I would love to disassemble the windows driver, but I have no experience 
> doing this. I was hoping to tap the collective experience of the group 
> before I tried that.
> 
> Any help would be much appreciated.
> 
>                                            Mboso
> 

Have you thought about to try to join the rt88xx backend people.
There will be differences between the scanner chips but maybe 
not so great.

http://hp44x0backend.sourceforge.net/

If you compare some usb-sniff output some parts can be the same?


-- 
--------
m.vr.gr.
Gerard Klaver