Imagine a clean white sheet of paper with years marked along the bottom. Now
imagine a line extending from the left side to the right side, over the edge
of the paper, across the wall, out the window and into the next town. We can
call this "MakeMaker's projected maintenance lifetime".
Now imagine me next to it, sobbing.
Actually it's not all that bad, I just like the image.
Point is, the hopes I had years ago that Module::Build or something would come
along and effectively kill MakeMaker aren't coming true. Uptake has been
slow. So it's going to have to soldier on. That said, I encourage new
projects to use Module::Build and existing projects to switch over. If you're
doing simple things, the switch will be simple. If you're doing complicated
things, it will probably be much more straightforward in Module::Build.
The basic goals of MakeMaker remain the same.
1) Keep it working on all platforms perl does.
2) Make it easier to maintain and test.
3) Support CPAN conventions and protocols (such as META.yml).
4) Transfer technology to other projects.
The "no new features" rule remains in place, they're just too damn hard to
test completely. But it is sometimes countermanded by goal 3.
An example of #4 was splitting out of ExtUtils::Install, Command and Manifest
into their own projects. This was fairly easy to do as they were already
separate modules. Other more involved ideas are to mine MakeMaker for its
extensive nuggets of portability logic and put them into some sort of separate
project for all to use.
--
ROCKS FALL! EVERYONE DIES!
http://www.somethingpositive.net/sp05032002.shtml
