Liane Praza writes: > We're getting to a world where creating distributions is much easier. I > hope that ARC will spend its time classifying interfaces and components > so that administrators, developers, users and distro builders alike can > make educated decisions, rather than having ARC attempt to define the > exact contents of a distribution.
We actually run into conflicting goals when we do this. On the one hand, we are trying to build a coherent system, which in part means that if you need to do "X," then there ought to be some small number of good ways to do "X" (as near to 'one' as we can get), and not a large number of crummy ways. On the other hand, if people deliver attractive-looking features, but everyone believes himself to be at the leaf of the tree -- everyone thinks it's fine if _his_ stuff is Volatile -- then we can't advise anyone to build on top of the system. It's sand thrown in the machine, just like all those private networking and file system interfaces that people have complained about for years. Neither ARC nor users nor distribution builders are in a good place to resolve issues like those. They're problems that have to be pushed upstream, to the designers of the software that ends up in the system. > John, I think the rest of your mail is a good way to start attacking the > problem. But I'm very wary of the idea that banishing non-ideal > software to another location solves anything. Focusing on what ARC > *can* do to classify seems like a much better approach. (Especially > when those classifications can be categorized and fast-pathed so we > don't see the same discussion in every ARC case.) I strongly agree with that. I think having separate "universes" or categories of non-compliant software just leads to a confusing mess. At a trivial level, there's the problem of choosing what goes in which universe. The answers are far from clear, and they're in part dictated by future projects we've never seen -- someone building on a component will want it to be "core" no matter where it actually is. More deeply, it's the very rare and trivial project that is just a leaf node in the architectural world. In the real world, it's often the case that seemingly leaf-like projects (say, CIFS or zoo) depend on changes in core-like bits (ZFS or GNOME), or vice-versa. And it's exactly those entanglements that the ARC is supposed to be reviewing. By pushing some subset of projects out into the "we don't care; do whatever what you want" portion of the universe, we haven't solved the problem. We've just created new ones. -- James Carlson, Solaris Networking <james.d.carlson at sun.com> Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084 MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
