[sane-devel] Canon imageCLASS MF5770 does not work with sane

[Jerry Zhang]

Hi Nico and Dennis,

I replaced the 2 files (pixma_imageclass.c & pixma_mp750.c) with the
new ones you provided.
Attached is the new log of "scanimage -T" after recompile/reinstall.

I sent you a large file (4MB, SnoopyPro capture of MF Toolbox) in my
previous e-mail.

Thanks,

Jerry


On Thu, May 14, 2009 at 2:30 PM, Nicolas Martin
<nicolas0martin at gmail.com> wrote:
> Hmm, this MF5770 holds a CCD device, like the MP750, that's why it
> requires a calibration sequence.
>
> Yes, a first step would be to try with the current MP750 backend file,
> and see how it goes.
>
> Jerry:
>
> Could you test the attached 2 files, simply replace the ones in your
> current git tree, in subdir sane-backends/backend with those 2 new ones
> in the archive, and recompile/reinstall
> In this case, take a logtrace as you did previously, and send it back.
>
> Also, as Dennis suggested, could you take a longer USB snoop to get the
> whole scan sequence ?
>
> Nicolas
>
>
>
> Le mercredi 13 mai 2009 à 18:16 -0700, Dennis Lou a écrit :
>> --- On Wed, 5/13/09, Nicolas Martin <nicolas0martin at gmail.com> wrote:
>> > Dunno if you still own your ImageClass device, but as you
>> > were more
>> > expert than me on this part of the pixma backend, I
>> > naturally dare
>> > proposing you, if you can, to have a quick look at what's
>> > going on
>> > here.
>>
>> Yup, still have it and I happened to catch the exchange on the mailing
>> list archive when I was browsing it.  Bad timing too, as I had just
>> cleaned out my old imageclass dev files and started a couple other
>> hacking activities a few days ago :-)
>>
>> Anyway, I took a quick peek at the provided log.  First of all, it's a
>> bit too short.  A little longer log might be helpful.
>>
>> Secondly, the protocol looks a lot more like the mp750 based on:
>> - Calibration cmd (0xe920) is in the logs like mp730/mp750 but not
>> iclass
>> - cmd_read_image length (10) is same as mp730/mp750 and unlike iclass
>> (11)
>> - cmd_err_info (0xff20) is in mp750 and iclass but not mp730
>>
>> Perhaps try taking the USB ID out of pixma_imageclass.c and put it in
>> pixma_mp750.c?  I can keep digging into the log if you want and
>> hopefully Jerry can supply a better one, but that's my initial
>> impression.
>>
>> -Dennis
>>
>> P.S. I find it a bit curious that a laser-based MFP would need
>> calibration.  Aren't these things supposed to be quick-start for when
>> the user wants to use it as a copier machine?
>>
>>
>>
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test1-0514.log
Type: application/octet-stream
Size: 80705 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20090514/8da1c52f/attachment-0001.obj>