[sane-devel] support for Canon MG8150
ralph.gauges at bioquant.uni-heidelberg.de
Tue Nov 2 09:23:45 UTC 2010
On 11/2/10 12:21 AM, Gernot Hassenpflug wrote:
> On Tue, Nov 2, 2010 at 5:29 AM, Louis Lagendijk
> <louis at lagendijk.xs4all.nl> wrote:
>> On Sun, 2010-10-31 at 19:51 +0100,
>>>> You can try adding the definitions near the end of pixma_mp150.c where
>>>> similar definitions exist, like this:
>>>> /* Latest devices (2010) Generation 4 CIS/CCD */
>>>> DEVICE ("Canon PIXMA MP280", "MP280", MP280_PID, 1200, 638, 877,
>>>> DEVICE ("Canon PIXMA MG5100", "MG5100", MG5100_PID, 1200, 638, 877,
>>>> DEVICE ("Canon PIXMA MG6100", "MG6100", MG6100_PID, 2400, 638, 877,
>>>> DEVICE ("Canon PIXMA MG6200", "MG6200", MG6200_PID, 2400, 638, 877,
>>>> DEVICE ("Canon PIXMA MG8100", "MG8100", MG8100_PID, 4800, 638, 877,
>>>> PIXMA_CAP_CCD | PIXMA_CAP_TPU),
>>>> and the USB ids near the start of the file under the existing ones,
>>>> like this:
>>>> #define MP280_PID 0x1746
>>>> #define MG5100_PID 0x1748
>>>> #define MG5200_PID 0x1749
>>>> #define MG6100_PID 0x174a
>>>> #define MG8100_PID 0x174b
>> Please check the USB-ids, if it is incorrect network scanning will work
>> (as long as the device id is unique) as the network code gets the
>> device-id from the definition in mp150.c where it does the lookup by the
>> name of the device read over the ethernet connection. For USB it works
>> the other way round....SO if the the id is incorrect network scanning
>> will work, while usb scanning fails
>> best regards, Louis (author of the BJNP stuff)
> Hi, yes, please confirm the USBid, I took the above ones from the
> Japanese MG8100 series driver, but do not have the device myself.
> Best regards,
> Gernot Hassenpflug
Yes, you are both right, I had made a slight mistake in the USB ids.
Now the scanner is recognized over ethernet as well as over USB. The
difference is that I can actually scan over ethernet, but get an error
when I try over USB.
I sent the debug output I get to Gernot and we are currently trying some
things to see why scanning over USB does not seem to work yet. Since I
was not sure whether to post all those log files on the list was OK, we
took part of the discussion offline.
Gernot suspected that the USB problems might be related to my version of
libusb. I was using the latest Ubuntu 10.10 AMD64 and the libusb with
the development packages that come with the system (0.1.12).
In order to be compatible with the system Gernot uses, I installed a
Debian Sid VM. Gernot also suggested to compile sane aginst libusb 1.0.8
instead of libusb 0.1.12, but so far I haven't succeeded in doing that.
configure always tells me that it didn't find libusb and that it will
compile without USB support.
Is there any trick to compile sane against libusb 1.0?
I won't be available until next week, but next week, I have planned to
get my hands on a Windows XP machine so that I can provide some info
acquired via USBSniffer
More information about the sane-devel