I've been reading this thread with some interest. Bill, I agree that if you don't want to try M::B and you're happy with EU::MM then by all means stick with it. But I do want to defend M::B's reason for existence.
It's true IMO that the main problem with EU::MM is its reliance on 'make', but not for the reasons already discussed in this thread. I [mostly] agree that platform differences in make dialects aren't really a big deal. The biggest problem with EU::MM is its whole paradigm: it has to squeeze all its functionality and abstractions through the little pigeonhole of Makefile syntax. That has lots of yucky implications, e.g. customization becomes difficult (if not downright dangerous), all actions must be performed via shell commands (which *do* vary wildly in syntax on different platforms), and trying to understand the dependency graph of the Makefile, which IMO is what Makefiles are *for*, is an exercise in headbanging. I'm sure it must have seemed like a good idea at the time EU::MM was written[1] to use 'make' for building Perl distributions, but IMO the bang we get for it is way less than the buck it takes from us. -Ken [1] Actually, it wasn't so much "written" as extracted through the nose of various scripts & misc. other code that was hanging around during the transition from perl4 to perl5.
