At 12:24 AM +1030 12/22/05, Tim wrote: >Tim: >>> When this, and system-config-bind updated, I found this problem: >>> >>> ll /etc/named.conf* >>> lrwxrwxrwx 1 root root 32 Dec 20 16:15 /etc/named.conf -> >>>/var/named/chroot/etc/named.conf >>> lrwxrwxrwx 1 root root 32 Oct 27 12:36 /etc/named.conf.rpmsave -> >>>/var/named/chroot/etc/named.conf >>> >>> Shouldn't the last one link to "named.conf.rpmsave"? For a moment I >>> thought I'd lost my customised version of named.conf. Luckily it was >>> still there, just not where I expected to find it. > >Tony Nelson: > >> Well, the .rpmsave files are just made by mv'ing the old file out of >> the way so it doesn't get clobbered. If a "file" is really a soft >> link, it will still point to the same path it used to point to, as it >> hasn't been changed. What you have with the .rpmsave files is a way >> to compare old and new, and a way to back out of difficulty if you >> need to, not some parallel working installation. > >I know what the RPMSAVE copies are for, I'm not sure if you noticed what >I listed. The /etc/named.conf.rpmsave file is really a link to the >chrooted named.conf file. It doesn't link to the chrooted >named.conf.rpmsave file. Of course I noticed what you listed. Read what I posted, and think about it. Look up any words you don't understand. Perhaps "mv" is one of them. >i.e. I reckon that it should have done things like this (below), not >like how it actually did it (above). > >ll /etc/named.conf* >lrwxrwxrwx 1 root root 32 Dec 20 16:15 /etc/named.conf -> >/var/named/chroot/etc/named.conf >lrwxrwxrwx 1 root root 32 Oct 27 12:36 /etc/named.conf.rpmsave -> >/var/named/chroot/etc/named.conf.rpmsave You "reckon" wrong. >Notice a slight difference? Yes. That difference was, in fact, the point of my post, which you should read carefully. Computers are not intelligent, or smart, or able to guess what you want for arbitrary data. >If they're going to bother to put symlinks in the /etc/ directory, they >ought to do so in a logical fashion. "They" didn't put any symlinks there, the Packager did. RPM did behave in a logical fashion. Your logic is wrong, and you are going to be fighting RPM and your computer until you stop disagreeing with it and learn about symlinks, mv, RPM, and packaging in general. ____________________________________________________________________ TonyN.:' <mailto:tonynelson@xxxxxxxxxxxxxxxxx> ' <http://www.georgeanelson.com/>