http://norman.walsh.name/2006/11/17/xprocwd 9 115 2006-11-17T12:44:36-05:00 $Date: 2007-04-05 09:55:02 -0400 (Thu, 05 Apr 2007) $ Norman Walsh 2006 Norman Walsh The XML Processing Model Working Group has published the second Working Draft of the pipeline language document. A few weeks ago, I asked the Working Group if they thought we could get another XProc Working Draft published sometime in December. They replied, why not next week? Is they great, or what? I think it actually took two weeks, but today, we published the second Working Draft of XProc: An XML Pipeline Language. There's still lots of work to be done, though I hear occasional, optimistic whispers that include “last call” and “January” in the same breath. I think the big changes are: The semantics have been rewritten so that they're declarative but without reference to a rather amorphous “flow graph”. The “declare-” forms of the input, output, and parameter elements are gone. The proponents of this change, myself among them, think this is actually simpler. Depending on the context, those statements are declaring, binding, or declaring-and-binding. Having two names for three concepts was, I think, more confusing than having one. The argument that if there are three concepts there should be three elements is compelling, but it's awfully verbose. If you have strong feelings about this (or anything else), please let us know. The source attribute no longer has two parts. (No microparsing!) Instead, there are two attributes, just like everyone said there should be last time I presented this work. Happy reviewing! (For more explanation and a little background, see my essay about our First Public Working Draft.)