See also: IRC log
Vojtech gives regrets for 1 Oct.
Vojtech: Close this one; latest editor's draft solves this one.
Vojtech: For the declared outputs, the spec outputs now says that it may be connected to the xpath context for some other step.
Norm: Ok. I tried to make that
clear, but failed. I'll expand on that.
... I'll enumerate all the cases.
Vojtech: Or if you can generalize
it somehow...I don't know.
... In the same section, an "output port may be connected to more than one input port."
... Could that be misleading because in a compound step it could be the output of a contained step?
... Maybe "an output port may have more than one binding"?
Vojtech: I'm not sure if binding is the right thing. We use both binding and being connected. I think sometimes they mean different things.
Norm: I'll make an attempt to finesse this somehow.
Vojtech: Is this section confusing wrt atomic and compound steps?
Norm: Yes, probably.
... I need to fix this.
Vojtech explains his concern.
Henry: A quick survey of the
likely specs didn't turn up a good definition, which surprised
me a bit.
... If "unordered collection which allows duplicates" is better...
... The problem is that any of these work, the question is which ones will work for which people.
... I can try harder to find a definition in the literature.
Norm: I don't know if "unordered
collection" carries the implication that duplicates are allowed
... I'd be just as happy to leave it, I think.
Some discussion of how to fix the markup
Henry will edit the XML and check it back in.
Proposal: Publish a new CR draft to satisfy our heartbeat requirements.
Norm: Henry, anything to work on this week?
Henry: No, no progress yet.