At 04:33 PM 12/22/98 -0500, Peter C. Norton wrote:
>For the heck of it, lets play out a scenario. It is similar to some
>that I've seen.
[snip]
>So, since you use rpm, you start by checking all of the recorded
>md5sums for the packages you've installed against the master database
>you have locked away on a day 1 backup tape. What's this! On every
I guess I'd have a more sophisticated comparison program. Why not have two
MD5 checksums - one for the data prior to the modified section of the binary
and one for after the modified section. Then I'd have a mechanism for
checking the modified bits. Not a pretty program, but not impossible either
I'd think.
Heck. Maybe the type of checking program should be tied to each file so that
you can vary between an MD5 check, a multi-part MD5 check, a diff (why not
for a text file that is never meant to change - more reliable than MD5) and
so on.
No, I'm not advocating anything, just observing that yes a simplistic MD5
has a problem, but why does it have to be only a simplistic MD5?
Regards.
