See also: IRC log
Skipping 27 Nov, US Thanksgiving.
No regrets heard.
Norm explains the situation and proposes to strike "or made available through p:namespaces"
Norm explains that it's only about user convenience.
Henry suggests that since XSLT 2.0 allows something similar, we should too.
Alex: I agree with Henry
Norm: If you match /, you get all the nodes in the document.
Mohamed: Is it the only place where we should do this?
Norm: After a quick check, I don't see any others that make any sense.
Proposal: Make the change.
Mohamed: What about www-form-url-encode.
Norm: Well...www-form-url-encode does allow match=/, but will invariably produce a dynamic error if you do that.
Norm: I think we should allow people to reuse the c: namespace.
Alex: I wouldn't want to restrict what people can produce from their steps.
Mohamed: We have three namespaces, on two of them we explicitly don't want people to reuse them.
Richard: What about in inline documents?
Norm: We're pretty clear that content in an inline is just content and we don't care what it is.
Mohamed: If we used the c: namespace and we use an element that's already been defined in this specification, can we add an attribute to it?
Henry: I think we should just say that common sense suggests that such usages shouldn't overlap with the uses defined in this spec. unless the usage is identical.
Alex: I'm with Henry.
Mohamed: Today we don't mandate that if you extend the output of an existing step, you must use an extension attribute. So you could put a @type on c:result and that might effect interoperability.
Norm: This is a bigger can of
worms than I thought, perhaps we shouldn't say anything.
... If we do this later, does it push us back in the process?
Henry: No, how could it effect the conformance of an implementation. This isn't a big deal.
<richard> X crashed :-(
Norm: So do we want to try to nail this down today, or come back to it later.
Henry: I'm fine to come back later
Mohamed: Me too.
Norm: Ok, we'll leave this.
Norm: Any comments?
Norm: Congrats to us all.
Mohamed: What's the future work?
Norm: We need to get the test suite finished, we need to encourage implementors, and we need to turn our attention to the default XML processing model.