No network scanner autodetection by backend kodakaio

[Paul Newall]

sane-backends 1.0,.23 in ubuntu 12.10 has a new backend kodakaio for kodak aio printers
The backend can autodetect network connected scanners using functions in cups.
The auto detect code is only compiled if cups.h is found at compile time.
It seems likely that cups.h was not found at compile time when the package was built?

If I compile and install kodakaio myself, the autodetect works with ubuntu 12.10
Can anyone confirm this diagnosis?
Is it possible to compile sane-backends with cups installed when building this package?
Or is it necessary for me to always compile the autodetect code and check for the presence of cups at run time?

[Paul Newall (p-newalls)]

Maybe there needs to be a dependency on libcups2-dev added to sane-backends ?

[Robert Ancell (robert-ancell)]

This will have to be done as a SRU for 12.10 but seems a valid request.

[Rolf Leggewie (r0lf)]

I don't see libcups2-dev among the build-deps, so I guess this is still a problem even in vivid?

[Paul Newall (p-newalls)]

Just for information. Later versions of sane-backends will auto detect the kodak AiOs using avahi instead of cups.
But there could be similar problems due to the compiled code depending on the presence/absence of avahi headers.

[Rolf Leggewie (r0lf)]

Thanks, Paul. Do you still own the scanner in question?

[Paul Newall (p-newalls)]

On 03/03/15 06:19, Rolf Leggewie wrote:
> Thanks, Paul. Do you still own the scanner in question?
>
Yes, I still have kodak aio printer scanners.
I'm not sure if working on this bug is still worth the effort.
Since sane backends 1.0.24 I think we are using avahi instead of cups to
do the network autodiscovery.
There is quite likely a similar problem, except with avahi instead of cups.

Paul

[Rolf Leggewie (r0lf)]

Paul, thank you for the reply. I am glad to hear you still have access to the hardware. I don't understand why you think working on this bug is still worth the effort when you believe that a similar problem is likely to exist in avahi. Can you elaborate? It would be great if you were able to confirm or refute that suspicion, too.

[Rolf Leggewie (r0lf)]

I don't understand why you think working on this bug is NOT worth the effort when you believe that a similar problem is likely to exist in avahi.

[Paul Newall (p-newalls)]

What I meant was that it was not worth working on autodetection using cups, when the method had changed to avahi.
I do not know if the avahi autodetection works for everyone, though I have not seen many complaints about that. But I guess I should be able to test it by seeing if the live DVD can autodetect a kodakaio printer.

[Paul Newall (p-newalls)]

I did a trial with a live DVD. Ubuntu 16.04
A network connected kodakaio printer can be auto detected. (I think that's with avahi)
scanimage -L lists the printer.

So this bug is solved.

[Rolf Leggewie (r0lf)]

Thanks, Paul, for sticking with us until the resolution of this issue.

Closing as resolved.