Re: attrchange

Ooops. Mea culpa; I forgot that the attrchange field was added as part of
the most recent tweak.  (More precisely, I forgot what name we had given
it, and I was looking at an outdated copy of the spec.)

It's a read-only field, so if users are ever going to generate instances of
this event it should indeed be settable in the init. One can argue that
users shouldn't be generating mutation events directly, but if we're going
to  support initMutationEvent() it probably should support this.

Good Catch, I think...

______________________________________
Joe Kesselman  / IBM Research