I'm curious about this, folks. Bear with me, please. In days of yore, all you ever heard was people whining about Windows NT (and, of course the DOS-based versions) just lying there while application and driver installations kicked the living daylights out of the OS by overwriting critical system files. I danced a jig when W2K's WFP features were announced.
Have you really run into many important apps or utilities that simply must have their way with the system file complement? W2K does have provisions for forcing an errant app to use "local" copies of system files (located in the program's own directory instead of in the usual system directory location). Of course it can be difficult to track down just which files you have to provide local copies of in order to accomplish this little trick, but IMHO it's well worth the time a trouble required to figure it out.
I was really tired of watching NT4 get hosed every time I had to add a device driver written by ATI or some other outfit that never got a clue about installation procedures, or whenever I tested some new "must-have" software seeing the system turn into a bowl of oatmeal. It's bad enough that installation procedures can still whack the registry so easily, and that's a much tougher issue for MS to deal with, too. I like my WFP. And I'd rather do without an app or a device if using means disabling WFP.
Am I missing something here? Is there a good technical reason for disabling WFP, or are our different approaches just the outcome of a difference in philosophies? (Me worship stability! Like stability! Stability good!)
Regards,
Jim