[sane-devel] Sane genesys Backend 1.0.24 CanoScan LiDE 210 and tiff scan format
Stef
stef.dev at free.fr
Tue Jun 18 19:15:46 UTC 2013
On 14/06/2013 20:30, Stef wrote:
> On 14/06/2013 17:46, Rolf Bensch wrote:
>> Hello,
>>
>> Am 13.06.2013 20:45, schrieb Stef:
>>> On 13/06/2013 12:26, Thorsten Müller wrote:
>>>> Hello,
>>>>
>>>> I've got a problem with the 1.0.24 genesys backend. I need to use
>>>> 1.0.24
>>>> since the 5 front panel buttons of my CanScan LidE 210 are not fully
>>>> supported in earlier versions.
>>>>
>>>> When I try to scan in tiff format, this it what happens:
>>>>
>>>> | tm at dungeonmaster:~ $ scanimage --format=tiff > /tmp/test.tiff
>>>> | scanimage: unrecognized option '--format=tiff'
>>>>
>>>> It works with earlier versions of sane (tested with 1.0.22 and
>>>> 1.0.23).
>>>>
>>>> I use Kubuntu Linux 12.04 Precise Pangolin x86_64 and the 1.0.24git
>>>> packages from this ppa:
>>>> http://ppa.launchpad.net/rolfbensch/sane-git/ubuntu
>>>>
>>>> scanimage --format=tiff works with the oroginal packages from the
>>>> distro
>>>> (Version 1.0.22-7ubuntu1) and from this ppa:
>>>> http://ppa.launchpad.net/nathan-renniewaldock/sane/ubuntu/ (Version
>>>> 1.0.23-0ubuntu1~ppa1~precise)
>>>>
>>>> I tried building from the daily 1.0.24 snapshot sources myself, it
>>>> gives
>>>> the same results:
>>>>
>>>> | tm at dungeonmaster:~ $ scanimage --format=tiff > /tmp/test.tiff
>>>> | scanimage: unrecognized option '--format=tiff'
>>>>
>>>> Perhaps someone can shed some light on this. And BTW, there are some
>>>> more long options that don't work any longer:
>>>> --icc-profile, --batch-*
>>>>
>>>> TIA for any help
>>>> Greetings
>>>> Thosten
>>>>
>>> Hello,
>>>
>>> I have ran into similar issue. The following patch fixes for me:
>>>
>>> diff --git a/frontend/scanimage.c b/frontend/scanimage.c
>>> index ce958df..49030a3 100644
>>> --- a/frontend/scanimage.c
>>> +++ b/frontend/scanimage.c
>>> @@ -2091,7 +2091,7 @@ Parameters are separated by a blank from
>>> single-character options (e.g.\n\
>>>
>>> /* re-run argument processing with backend-specific options
>>> included
>>> * this time, enable error printing and arg permutation */
>>> - optind = 0;
>>> + /* optind = 0; */
>>> opterr = 1;
>>> while ((ch = getopt_long (argc, argv, full_optstring,
>>> all_options,
>>> &index)) != EOF)
>>>
>>> Since the current has been correct for years, maybe there is a
>>> regression in glibc ?
>>>
>>> Regards,
>>> Stef
>>>
>>>
>> The Pixma backend is running with and without this patch on my 64-bit
>> Ubuntu 12.04. I have no problems with any linux kernel (tested with
>> kernel version 3.2.0 and 3.8.0) without this patch.
>>
>> 'cause my ppa is affected, I would like to have an official patch for
>> this issue.
>>
>> Somebody who is involved in coding scanimage should tell something about
>> this. I have no idea about the usage/need of the optind parameter in the
>> getopt library.
>>
>> Cheers,
>> Rolf
>>
>>
> Hello,
>
> for me the scanimage code is perfectly correct. I don't think
> anything needs fixing in it. My "patch" is a workaround for a bug that
> comes from elsewhere.
> So the ppa you built is working fine for you and failing for others ?
> Are glibc used different, or compiled with a different compiler ?
>
> Regards,
> Stef
>
Hello,
I eventually found the root of the problem. Two option groups
didn't have there .cap value defined explicitly. This was working until
it broke recently, possibly related to the use of a newer compiler.
However I pushed a fix in git, and now this issue is fixed.
Regards,
Stef
More information about the sane-devel
mailing list