On Tue, Nov 5, 2013 at 7:35 PM, Yasha Karant <ykar...@csusb.edu> wrote: > I have not tried this yet, but the statement is that "The cmp utility > compares two files of any type and writes the results to the standard > output." > > Will this work for mount points or device special files that are mounted on > mount points, allowing one to compare two full directory tree hierarchies?
Yup! The idea that "a file is a file, even when it's a mountpoint or a symlink or a pipe or a socket" is fundamental to the concept of what a "file" is in libc and in the kernel. itself. > That is, using the situation elaborated below (that is, earlier in this > thread), will > > cmp /dev/loop0 /dev/sr0 Should work, if those are the valid device names. If you can 'md5sum' it, you can 'cmp' it. > compare the ISO image and the original mounted DVD (after driver error > correction, if any) "byte by byte"? > > On a separate issue, when burning a DL DVD-ROM (single sided, approximately > 8 Gbyte storage), is there a difference between a DVD+R and DVD-R in the > final physical product? If so, is there a mechanism (software) to determine > if the source DVD being cloned (copied) is -R or +R ? All of the mechanisms > I have found (to date) report the source DVD as a DVD-ROM, with no > indications of +R or -R. Interesting question. I wouldn't expect the cloning process to replicate that data, offhand, but I'd expect the cloning tool itself to be able to report it when the clone is being made. This sort of thing is fundamental to DRM, where the provenance of the media is considered critical. The backchannel information that identifies the manufacturer of the DVD or CD to provide the information for the drive itself to know the specific information needed to read or write the media might be accessible to the CD or DVD drive's firmware. That's why virtualization of the guest operating systems, permitted to mount hardware accessible to the virtualization hosting service, has already effectively broken so-called "Trusted Computing" based DRM.