In the latest spec file: # Remove permissions from udev rules, but register for ACL management examples/hotplug -a"SYMLINK+=\"libmtp-%k\", ENV{ACL_MANAGE}=\"1\"" > libmtp.rules Apparently ACL_MANAGE shouldn't be used in udev rules anymore, and broke gnome-bluetooth (see bug 588660 and the links near the end).
OK edited the Fedora extenstion to remove this, built to RawHide, is it working out?
Are you sure that just removing the ACL_MANAGE is sufficient? I think you'll have to set the TAG+="udev-acl" instead, as e.g. http://svn.debian.org/wsvn/pkg-gnome/desktop/unstable/gnome-bluetooth/debian/61-gnome-bluetooth-rfkill.rules or http://cvs.fedoraproject.org/viewvc/rpms/gnome-bluetooth/devel/61-gnome-bluetooth-rfkill.rules?revision=1.2&view=markup does.
Just removing ACL_MANAGE isn't enough, and apparently adding udev-acl is frowned upon. You'll need to ask Harald or upstream udev what to do in this case.
OK now it says ENV{ACL_MANAGE}="0" TAG+="udev-acl" But if adding udev-acl is frowned upon, then it's not good for RFKill either I guess? This is more than a bit idiomatic, is there some post I can read up on the general idea of these strings and how it's supposed to work these days...
ENV{ACL_MANAGE}="0" is incorrect. ENV{ACL_MANAGE}!="0" as an added condition would make sense.
(ENV{ACL_MANAGE}="0" is an assignment, not a condition, ENV{ACL_MANAGE}=="0" would be a condition, but the wrong one: if you check the gnome-bluetooth rules carefully, it SKIPS the block if ENV{ACL_MANAGE}=="0".)
This bug appears to have been reported against 'rawhide' during the Fedora 14 development cycle. Changing version to '14'. More information and reason for this action is here: http://fedoraproject.org/wiki/BugZappers/HouseKeeping
Well I don't know how to proceed with this because I just don't know what to do. I obviously know too little about udev to make use of the advice given here, please help. A patch is very good help, for example...
With the current udev in F14, ACL_MANAGE is not used anymore at all, so just TAG+="udev-acl" is likely to work. But as far as I understand, a proper long-term solution would be for the udev authors to add some ENV{ID_*} variable for each library which needs one so that libraries don't have to set that udev-acl tag themselves.
This is fixed in rawhide now, closing.