> XML in C++ has been lacking a language binding for a long, time, > and (I believe) W3C has no intention of providing one.
I'd say "no current plan" rather than "no intention". I think the DOM WG's position has been that we'd like to see more standardized language bindings, but we'd like to see the language communities come up with a proposal rather than our spending cycles on it. The right thing is probably for folks to look at the existing bindings for other languages, synthesize one that includes all their best features, and submit it to W3C for publication as a NOTE; if that holds up under examination by other folks, it might eventually migrate from NOTE status into being an official part of the DOM spec. Note that bindings may have to reflect both language and linkage mechanism; the best C++ DOM binding for use thru Corba might not be identical to the best choice for local manipulation. Also, since the DOM explicitly doesn't address memory management issues, whoever develops these bindings has to deal with whether they're assuming automatic recovery of disused objects or expect the user to issue a release/discard/free operation of some sort. ______________________________________ Joe Kesselman / IBM Research
