> Good job on this.
Thanks :)
And it's now on the list in case someone runs into the same problem somehow.
Simon
>
> On 1/24/26 10:50 AM, Simon Matter via sane-devel wrote:
>>> Hi,
>>>
>>> Would it be possible for someone to look at the button daemon and see
>>> if
>>> there are obvious errors that you lead to issues as discussed here.
>>>
>>> The code is attached again to this mail.
>> Finally I found out how this can be solved. The following patch does the
>> trick:
>>
>> @@ -59,7 +59,11 @@ int
>> main (int argc, char **argv)
>> {
>> int i, idx;
>> - struct timespec ts = { .tv_sec = 0, .tv_nsec = 100000000L };
>> + /* Important query rate limiting, the old value of 0.1s was too
>> + * small. USB scanners can block on newer Linux systems if the
>> + * query rate is too high.
>> + */
>> + struct timespec ts = { .tv_sec = 0, .tv_nsec = 500000000L };
>> time_t last_reopen;
>> SANE_Bool new, print_err;
>> const SANE_Option_Descriptor *opt[MAX_BUTTONS];
>>
>> Thank God this is solved now before I gave up on it! Still, I believe
>> this
>> is a bug in recent Linux systems and shouldn't happen. But I tested from
>> MX Linux to latest Fedora so many systems and it's always there.
>>
>> Regards,
>> Simon
>>
>>> Thanks,
>>> Simon
>>>
>>>> I start to believe the problem is not in SANE itself but in the button
>>>> daemon we are using (see attachement).
>>>>
>>>> Our software is running the button daemon and waits for events, then
>>>> terminates the button daemon and does the appropriate scanimage job,
>>>> and
>>>> then starts the button daemon again.
>>>>
>>>> The button daemon is a modified version of some daemon which was
>>>> floating
>>>> around many years ago. It has worked well for the last 15 years or so.
>>>> Maybe there is some modification needed to make it work well with
>>>> newer
>>>> SANE.
>>>>
>>>> If someone who knows SANE better than me would take a look at it, that
>>>> would be much appreciated.
>>>>
>>>> Thanks,
>>>> Simon
>>>>
>>>>> We need to isolate the problem if possible- I expect if you compile
>>>>> an
>>>>> older, known working sane-backends on your current computer, the
>>>>> problem
>>>>> will persist. That would indicate hardware or OS level issue.
>>>>>
>>>>> allan
>>>>>
>>>>> On Wed, Jan 14, 2026 at 9:35 AM Simon Matter via sane-devel <
>>>>> [email protected]> wrote:
>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> On 2026-01-14 14:59, Simon Matter via sane-devel wrote:
>>>>>>>> These scanners work perfectly fine with very old sane-backends
>>>>>>>> and operating system.
>>>>>>> Does it work on the exact same computer
>>>>>>> where it fails or does it work on old computers?
>>>>>> It works fine on old computers with old software.
>>>>>>
>>>>>>> I ask because of possible USB 3 related issues, see
>>>>>>> https://en.opensuse.org/SDB:Configuring_Scanners#USB
>>>>>> Thanks, I'll check it out.
>>>>>>
>>>>>> Simon
>>>>>>
>>>>>>
>>>>>>
>>>>> --
>>>>> "well, I stand up next to a mountain- and I chop it down with the
>>>>> edge
>>>>> of
>>>>> my hand"
>>>>>
>>
>>
>
>
>
