Bug#382195: Offer of Assistance
* Sean Kellogg <skellogg@gmail.com> [2007-02-01 09:12]:
> Not sure what to do about the man pages. These tools are all very tiny and
> not exactly used for much other than testing your libmtp install. What if
> one man page were written with just a very general explenation of how to use
> the tools and then use that as the man page for all of them? I feel like
> this is done with other packages that have a bunch of tiny binaries.
I uploaded a new version of the package to my usual apt-getable repository
[1]. There is now a generic man page called mtp-tools.1, which all the
other man pages link to.
Jean Parpaillon offered me his work, so I am "adopting" the package. I will
upload it to unstable soon, unless there are big objections. In particular,
if you have suggestions on how to improve the packages descriptions, I will
be grateful.
> Is audio the right group? I've alwasy been confused by how udev OUGHT to
> work, so I'll defer to experts, but it seems to me that having access to
> the audio device on the system is different than having the ability to
> write to a usb device. I set mine to plugdev, only because it seems to
> have to do with people who 'plug' things in. But, like I said, I have no
> idea how this is supposed to work. Perhaps the policy says something?
I guess that the Policy mention this. The base-passwd package has a
document [2] which should be authoritative on these matters. It reads:
===========================================================================
audio
This group can be used locally to give a set of users access to an audio
device.
[...]
plugdev
Members of this group can mount removable devices in limited ways via
pmount without a matching entry in /etc/fstab. This is useful for local
users who expect to be able to insert and use CDs, USB drives, and so
on.
Since pmount always mounts with the nodev and nosuid options and applies
other checks, this group is not intended to be root-equivalent in the
ways that the ability to mount filesystems might ordinarily allow.
Implementors of semantics involving this group should be careful not to
allow root-equivalence.
===========================================================================
Neither of the above groups suit perfectly the MTP devices. I would happily
use any of them in the libmtp package.
What do the others think?
[N.B.: If you are in the Cc: list and do not wish to receive the followups,
please tell me.]
--
Rafael
[1] http://people.debian.org/~rafael/libmtp/
[2] /usr/share/doc/base-passwd/users-and-groups.txt.gz
Reply to: