[sane-devel] Invalid argument (status 4) when calling sane_start

2007-12-16 Thread Peter Gebauer
Hi!

 The v4l backend supports only a few modes so you have to select the correct 
 ones to get a image.

So it's the v4l settings that are wrong or the SANE settings?

There's a couple of things I'd fix with the v4l backend

1) Default (startup) mode must be valid and supported so a user can 
sane_start() with the default settings. It makes no sense to put the device 
in broken mode at startup.

2) Unsupported modes should not be listed. Why whould you even list 
options/modes that aren't supported by the backend? Nobody wants a 
SANE_STATUS_INVAL so nobody will set an unsupported mode anyway. Any modes 
not supported should be removed from the options.

If your sane_control_option didn't give an error then that mode should be 
considered valid and usable setting. This will remove the unwanted surprise 
when calling sane_start() and it will remove the guess work trying to figure 
out what option is breaking the sane_start() call.

I'll look at the v4l backend, but fixing it to work properly would be a 
major detour from my project. The idea of allowing SANE to use v4l devices 
is great, it feels generic and... cool. :)

/Peter




[sane-devel] Invalid argument (status 4) when calling sane_start

2007-12-16 Thread m. allan noah
On 12/16/07, Peter Gebauer petgeb at users.sourceforge.net wrote:
 Hi!

  The v4l backend supports only a few modes so you have to select the correct 
  ones to get a image.

 So it's the v4l settings that are wrong or the SANE settings?

 There's a couple of things I'd fix with the v4l backend

 1) Default (startup) mode must be valid and supported so a user can
 sane_start() with the default settings. It makes no sense to put the device
 in broken mode at startup.

 2) Unsupported modes should not be listed. Why whould you even list
 options/modes that aren't supported by the backend? Nobody wants a
 SANE_STATUS_INVAL so nobody will set an unsupported mode anyway. Any modes
 not supported should be removed from the options.

 If your sane_control_option didn't give an error then that mode should be
 considered valid and usable setting. This will remove the unwanted surprise
 when calling sane_start() and it will remove the guess work trying to figure
 out what option is breaking the sane_start() call.

 I'll look at the v4l backend, but fixing it to work properly would be a
 major detour from my project. The idea of allowing SANE to use v4l devices
 is great, it feels generic and... cool. :)

yes, it does sound nice, but sane devels are generally volunteers, who
only have time to scratch their particular itch. the v4l backend is
clearly labled as ALPHA in its man page. Since the rest of us dont
seem to use it, maybe it was waiting on you to finish it :)

allan


 /Peter


 --
 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



[sane-devel] Invalid argument (status 4) when calling sane_start

2007-12-16 Thread Gerard Klaver
On Sun, 2007-12-16 at 10:32 +0100, Peter Gebauer wrote:
 Hi!
 
  The v4l backend supports only a few modes so you have to select the correct 
  ones to get a image.
 
 So it's the v4l settings that are wrong or the SANE settings?
 
 There's a couple of things I'd fix with the v4l backend
 
 1) Default (startup) mode must be valid and supported so a user can 
 sane_start() with the default settings. It makes no sense to put the device 
 in broken mode at startup.
 
 2) Unsupported modes should not be listed. Why whould you even list 
 options/modes that aren't supported by the backend? Nobody wants a 
 SANE_STATUS_INVAL so nobody will set an unsupported mode anyway. Any modes 
 not supported should be removed from the options.
 
 If your sane_control_option didn't give an error then that mode should be 
 considered valid and usable setting. This will remove the unwanted surprise 
 when calling sane_start() and it will remove the guess work trying to figure 
 out what option is breaking the sane_start() call.
 
 I'll look at the v4l backend, but fixing it to work properly would be a 
 major detour from my project. The idea of allowing SANE to use v4l devices 
 is great, it feels generic and... cool. :)
 
 /Peter
 
 

The v4l backend doesn't support all v4l modes.
See also SANE bugreports for some update information (30015,30016,30018)

-- 

m.vr.gr.
Gerard Klaver