Hi Marco.
On Fri, May 14, 2010 at 01:19:53PM +0200, Marco d'Itri wrote:
> Changing $udev_root has never been properly tested or documented,
> considering the recent focus on devtmpfs I do not believe that it will
> continue to work for a long time no matter how many hacks I could add.
i respect Your expertise and Your efforts to keep udev running.
anyway what you claim here is plain wrong. let me explain:
- the testing of a non-default value for $udev_root has taken place.
i can at least speak for myself, having changed this setting shortly
after i encountered udev. as i wansn't involved in fixing the
/etc/init.d/udev{,-mtab} scripts i conclude that theres somebody else
caring about $udev_root.
- in fact $udev_root is documented. don't You ship the manpage?
i confirm that it is working as described in this manpage.
- the move from /dev to $rootmnt$udev_root does hardly change when
switching from tmpfs to devtmpfs. this will continue to work whatever
filesystem will be used for udev in furure releases.
- You mistook my work as a 'hack'. this is the cleanest possible way to
get the initramfs-tool make its job and leave behind a system as the
user wishes -- just like yaird does (did with kernels <= 2.6.30)
to fix this later on (after chroot) would be a serious hack.
- udev_root does not occur in init-top/udev or init-bottom/udev scripts.
so obviously this is not about how many hacks are involved here.
could You explain why You are opposed to allow the user to choose for
himself? If there's already a discussion on this, please redirect me.
Your strategy about "removing every trace of $udev_root" doesnt sound
amusing.
I'd really hope You take some time and rethink this point.
regards
felix
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
