On Fri, 06 Sep 2002 11:01:36 +0000, H. Narfi Stefansson wrote: > On Friday 06 September 2002 10:53, Frederic Crozat wrote: >> On Fri, 06 Sep 2002 16:46:55 +0100, Alastair Scott wrote: >> > --=-uv4Dj+q907OgonNFRe5O >> > Content-Type: text/plain >> > Content-Transfer-Encoding: quoted-printable >> > >> > On Fri, 2002-09-06 at 14:25, H. Narfi Stefansson wrote: >> >> I thought the whole idea was that this should be automated now. Does >> >> the= >> > >> > =20 >> > >> >> 'automated' part only apply to the mounting and not to the >> >> unmounting? In other words, is this a feature or is this a bug? >> > >> > It looks like a bug, and I reported it during the 'great mail server >> > outage' but it appears that that post never got through. So ... >> > >> > I have a Sony Clie PEG-425T and used to mount its memory stick 'by >> > hand' from the command line as /dev/sda1.=20 >> > >> > With the automated facility, under Gnome I see a 'removable' icon on >> > the desktop (with the rather inappropriate appearance of a hard disk!) >> > but: >> > >> > i. the icon appears to be stuck there forever and never goes away, >> > even when the Clie is taken out of the cradle and the PC rebooted >> > without it; >> >> Indeed.. Because it is using supermount.. No problem here.. >> >> > ii. the 'Unmount Volume' command has no effect. >> >> In supermount mode, "Unmount Volume" doesn't do anything since you are >> not root.. I'll try to hide it (no garantee) >> >> > iii. the entry in /etc/fstab looks like >> > >> > none /mnt/removable supermount >> > dev=3D/dev/sda1,fs=3Dauto,--,iocharset=3Diso8859-15,codepage=3D850,uma >> >sk=3D= 0 0 0 >> > >> > This looks OK, but doesn't explain the unmounting not happening: I >> > have to issue umount -a /dev/sda1 from the command line to actually do >> > it. >> >> There is no unmounting because usb-storage doesn't remove the /dev/sda1 >> entry when clié is no longer connected.. Therefore, there is no way to >> know the device is here or not :(( > > Can this be changed in the usb-storage module or is nobody willing to > touch the kernel modules issues at this point, or is it simply too > difficult?
First, don't CC: people when responding to mailing list. Second, I don't know but I thing we won't touch kernel to fix this minor issue since chances would be high to break it somehow.. -- Frederic Crozat MandrakeSoft
