At 01:49 AM 3/1/00, Ian Turner wrote: >Which is why effective quota security should enable inode limits as well >as byte limits. If I can take up all the useable clusters with 0-byte >files, that is just as bad as being able to take up the useable space >with >1-k files. NT4 had no quotas so the complaints were few, although there was a call for them. Now MS put them in and we are assuming that their purpose is to prevent DoS attacks. I've worked with systems with disk quotas for more decades than I'd like to admit and we never looked at them as a way to prevent malicious people from filling up the disk. Their main purpose was historically to prevent careless or greedy users from tying up space by forcing them to maintain their on line storage. Quotas worked well for this purpose. If this is the philosophy behind the MS implementation, it will do its intended job just fine. The fact that it could have solved another problem as well may make the implementation fall short of our desires, but that doesn't make it buggy (IMHO), just not what we, with 20-20 hindsight, would like to see them have done. Maybe they'll improve it if we ask nicely. There are lots of other things that MS does that I'd like fixed that are higher on my priority list. Joe ------------------------------------ There is always an easy solution to every human problem -- neat, plausible, and wrong. -H. L. Mencken
This archive was generated by hypermail 2b30 : Fri Apr 13 2001 - 15:38:40 PDT