Re: SANE and saned

David Mosberger-Tang (David.Mosberger@acm.org)
Tue, 21 Jul 1998 21:58:56 -0700

>>>>> On Mon, 13 Jul 1998 15:09:47 +0100 (BST), Ian Grant <I.A.N.Grant@damtp.cam.ac.uk> said:

Ian> I'm running a tcp-wrappered saned on a linux box and accessing
Ian> it from Digital Unix machines and Intel Linux machines across
Ian> the network. It works well, except I don't understand how the
Ian> saned.conf file works. I have no saned.conf file anywhere and
Ian> saned seems to talk to anyone the tcp wrappers let through.

Is it possible that all hosts are listed in /etc/hosts.equiv? As per
saned man-page:

FILES
/etc/hosts.equiv
The hosts listed in this file are permitted to
access all local SANE devices. Caveat: this file
imposes serious security risks and its use is not
recommended.

Note: I'd recommend against using /etc/hosts.equiv for security
reasons, but if a host is listed in /etc/hosts.equiv already, there is
little point in saned denying access to such a host.

Ian> Also, the saned man page claims that saned and the net backend
Ian> use service 'sane' when in fact they look for 'saned'.

Oops, this is indeed a bug in saned. The next version of SANE will
have this corrected. Unfortunately, this will break existing
installations, but the service name definitely shouldn't have a `d'
appended.

Thanks for finding this!

Ian> Once I've given the version of xscanimage and the net backend a
Ian> thorough testing on Digital Unix I'll send in a report and we
Ian> can declare 0.73 at least partly tested and found working on DU
Ian> 4.00 (I still want to try getting the SCSI/CAM on DU working
Ian> though. I'm waiting for someone to tell me what to do.)

Yes, working CAM support would be nice. I can't tell you what to do
other than "try it out and see what breaks". Be careful though when
playing with this as faulty SCSI code could harm other SCSI devices.

--david

--
Source code, list archive, and docs: http://www.mostang.com/sane/
To unsubscribe: echo unsubscribe sane-devel | mail majordomo@mostang.com