Re: mounting error USB stick
- To: firstname.lastname@example.org
- Subject: Re: mounting error USB stick
- From: Adam Hardy <email@example.com>
- Date: Mon, 03 Mar 2008 10:28:16 +0000
- Message-id: <[🔎] 47CBD2C0.firstname.lastname@example.org>
- In-reply-to: <20080229193824.GB27796@pc0197>
- References: <20080224230536.GA4777@pc0197> <47C20658.email@example.com> <20080225173900.GA17126@pc0197> <47C35A64.firstname.lastname@example.org> <20080226192859.GA31183@pc0197> <47C53512.email@example.com> <20080227220130.GA10282@pc0197> <47C68FA2.firstname.lastname@example.org> <20080228135309.GA7549@pc0197> <47C7E61D.email@example.com> <20080229193824.GB27796@pc0197>
Florian Kulzer on 29/02/08 19:38, wrote:
On Fri, Feb 29, 2008 at 11:01:49 +0000, Adam Hardy wrote:
Florian Kulzer on 28/02/08 13:53, wrote:
The sourceforge website is so slow! Anyway, ivman does the job, although
it mounts my usbstick as /media/sda1, so I'm not sure it will work with
more than one. I haven't got a second usb stick handy to test it with. It
would be nice if it mounted the usb stick using its volume label.
Since you want to be independent of the desktop environment, you might
want to run ivman, usbmount, or a similar daemon. If I remember
correctly, it is difficult to ensure unique mount points with usbmount
since it relies on plug-in order. I like the concept of ivman better
because it relies on HAL, so it should not clash with a running desktop
environment. I have never tried ivman myself, though. Its sourceforge
page says it is currently in beta stage. (Both usbmount and ivman are
available in Etch.)
I had hoped that this was one of the "user-configurable actions" that
are mentioned in the package description of ivman.
that is actually a trick of Thunar (or whichever package it was in xfce).
That is difficult to say without looking at the source code. You can
always specify a desired mount point policy, which is added to the
properties of the device in HAL's list. Furthermore, HAL does provide
certain methods for devices, for example "mount", "unmount" and "eject",
but it is easily possible to bypass these and (un)mount devices without
involving HAL directly. Maybe ivman only uses HAL to discover new
removable devices but does not care about the volume label.
But you say it is Hal using the volume label?
You can try to specify volume.policy.desired_mount_point explicitly for
all removable storage devices. If ivman uses HAL's abstracted "mount"
method then this should be honored (as far as I know).
More information about all this can be found here:
/usr/share/doc/hal-doc/spec/hal-spec.html (in the "hal-doc" package)
To get you started, here is an example of a HAL device information file
that requests the mount options "noatime" and "sync" for all partitions
on removable media (change this to reflect your own preferences).
Furthermore, the desired mount point is set to the volume label if the
label exists and if it contains only ASCII characters:
That sorted it out! Grand. Thanks alot for the info. Just FTR, hal doesn't have
an /etc/init.d start-up script. How it starts at boot is a mystery, so I just
killed it off and restarted it with hal --daemon=yes
I couldn't find the info in ivman about setting up the mount point using the
volume label, but since hal does it via its own configuration, I guess I don't
need to know now. The ivman docs referred to a 'default file provided with
ivman' containing 'many more examples' but I couldn't find it and I figured it
got lost from the ivman package.
The ivman forum has messages about development a new improved program to replace
ivman (called halevt), so I guess ivman isn't maintained that well.