[sane-devel] Observations whilst testing SaneTwain with Mustek 1200CP

jim.george@blueyonder.co.uk jim.george@blueyonder.co.uk
Tue, 15 Jan 2002 16:15:27 +0000 (GMT)


Henning,

	sorry it's taken so long to write back re-this subject but I've been
re-compiling from the CVS version to add in gphoto2 support (as you know ;)).

	Interestingly I've been attempting to use sanetwain remotely (from work)
and if I crash the client saned is now detecting it (saned received signal 13
exiting), so something improved there.

	I think I did run saned -d128 as root, however it now appears to work
again...so looks like operator error I'm afraid.

	Again sorry for the delay and also sorry that I had nothing more useful
to report.

	I have seen two occurences of saned 'apparently' taking out my pppd link
to the net.  But I can't confirm this. I'll monitor and let you know if this
happens again.

	Is it possible to have saned configured in inetd.conf and run saned -d128?
	I ask because it would make testing easier than having to disable saned
in inetd.conf everytime I want to test for potentially spurious events.

Thanks,

Jim
On Tue, 8 Jan 2002, Henning Meier-Geinitz wrote:

> Hi,
>
> On Mon, Jan 07, 2002 at 05:57:40PM -0000, jim.george@blueyonder.co.uk wrote:
> >         1) Although there is documentation related to the ports used by
> > saned, it is not prominent that saned uses 6566 AND other (seemingly)
> > random ports for the actual data processing.  This could be an issue for
> > those wishing to allow scanning through a firewall.  Is there a specified
> > number/range of ports? Could these be more prominently documented?
>
> I had only a quick look at this and it seems the ports are chosen
> randomly by the operating system. saned creates the socket and
> transmitts the port no to the net backend. I have added a TODO entry
> for documenting this issue. Maybe we can find a nicer solution for
> SANE 2.
>
> >         2) When the client dies unexpectedly saned doesn't appear to handle
> > this correctly, especially when run through inetd.  I have tried both
> > documented options re-inetd.
>
> Depends on what you mean by "dies unexpectedly". segfaults etc.
> shouldn't happen in the backends (-->bug). Could you write what you
> did and send a debug log (saned -d and maybe SANE_DEBUG_MUSTEK_PP=255)?
>
> >         4) When I ran saned -d128 I couldn't detect my scanner when
> > executing scanimage -L.
>
> Is this a problem of running it with -d or with the high debug level?
> Do you run saned -s as root (probably necessary for PP scanners)?
> Otherwise, a logfile would be nice.
>
> Bye,
>   Henning
> _______________________________________________
> Sane-devel mailing list
> Sane-devel@www.mostang.com
> http://www.mostang.com/mailman/listinfo/sane-devel
>