[sane-devel] canoscan 300 scsi scanner not detected

[m. allan noah]

Several comments:

1. the backend should be changed to look at multiple rows of its config file.

2. The canon backend code does this:

 if (ibuf[0] != 6
      || strncmp ((char *) (ibuf + 8), "CANON", 5) != 0
      || strncmp ((char *) (ibuf + 16), "IX-", 3) != 0)
    {
DBG (1, "attach: device doesn't look like a Canon scanner\n");
.....

So the apple machine must respond as a canon too.

3. The sanei_scsi code appears to use strncmp with the length of the
string from the config, so this should work:

scsi CANON IX

Denis- can you put only that line in your config file, and verify that
the scanner still appears?

allan

On Thu, Oct 28, 2010 at 10:35 AM, Johannes Meixner <jsmeix at suse.de> wrote:
>
> Hello,
>
> On Oct 28 09:52 m. allan noah wrote:
>>
>> I would need to check the code, but I think the canon backend mostly
>> supports machines who's model name starts with 'IX' so the best
>> solution might be:
>>
>> scsi CANON IX*
>>
>> or some such?
>
> I do not have a scanner which is supported by the "canon" backend
> so that all what I write here is only based upon what I read
> somewhere.
>
> According to what I read in "man sane-scsi"
> ---------------------------------------------------------------
>    scsi VENDOR MODEL TYPE BUS CHANNEL ID LUN
>
> ... a field's value can be replaced with an asterisk
> symbol (``*'').  An asterisk has the effect that any
> value is allowed for that particular field.
> ---------------------------------------------------------------
> so that it seems it is not possible to replace only a part
> of a particular field with an asterisk.
>
> According to
> http://www.sane-project.org/sane-backends.html#S-CANON
> the "canon" backend supports at least one non-CANON scanner,
> the "Apple Color OneScanner 600/27".
> I have no idea how this device shows up at the SCSI
> but I assume Apple made sure that "their device" shows up
> at the SCSI under a manufacturer name like "APPLE".
>
> According to my tests in
> http://lists.alioth.debian.org/pipermail/sane-devel/2006-May/016922.html
> it seems it is not possible to have something like
> -------------------------------------------------------------------
> #canon.conf
> scsi CANON
> scsi APPLE
> -------------------------------------------------------------------
> because it seems only the last entry "scsi APPLE" is used then.
>
> According to my tests in
> http://lists.alioth.debian.org/pipermail/sane-devel/2006-May/016922.html
> it seems not to cause problems if devices are probed
> which are no scanners so that in the end our patch results
> an unrestricted plain "scsi" by default in /etc/sane.d/canon.conf
> -------------------------------------------------------------------
> # canon.conf
> #
> # Activate exactly one of the following choices:
> #
> # Probe on all /dev/sg* devices:
> scsi
> #
> # Probe for "CANON" on all /dev/sg* devices:
> #scsi CANON
> #
> # Probe only on one fixed device:
> #/dev/sg0
> -------------------------------------------------------------------
>
> We have this since July 2006 and at least I noticed no bug reports
> or user complaints because of this - but I have no idea how many
> users use scanners which are supported by the "canon" backend.
>
> By default we have all backends in /etc/sane.d/dll.conf disabled
> (to avoid that a SANE frontend lets various backends start up
> and do needless probing for devices which do not exist)
> so that the "canon" backend only runs for those users which
> actually use a matching scanner.
>
>
> Kind Regards
> Johannes Meixner
> --
> SUSE LINUX Products GmbH, Maxfeldstrasse 5, 90409 Nuernberg, Germany
> AG Nuernberg, HRB 16746, GF: Markus Rex
>
> --
> sane-devel mailing list: sane-devel at lists.alioth.debian.org
> http://lists.alioth.debian.org/mailman/listinfo/sane-devel
> Unsubscribe: Send mail with subject "unsubscribe your_password"
>            to sane-devel-request at lists.alioth.debian.org
>



-- 
"The truth is an offense, but not a sin"