Meeting minutes
Binding Expression Constraint
https://
Steven: Not too sure what to do about this term
… or even what is implied
Steven: Do either of you dispatch an xforms-computer-error?
Alain: I'd need to check
Erik: I don't think so
Steven: It worries me that normative style text is being used in an informative section
… "This appendix gives an example of how a recalculate might be done. "
Erik: This problem has probably been there since 1.1
Erik: We could lower-case it...
Erik: Looking at <bind> we don't say anything about simple content for calculate and initial, and we should
ACTION: Steven to check that @initial and @calculate only refer to simple content
Steven: I'm tempted to lower-case it, so it doesn't look like a defined term, since it is non-normative anyway
Steven: So resolved
Error types
https://
My proposal is not to have an error-type for binding-errors at all
Erik: Sure
Alain: Yes
<output mediatype=""/>
https://
Steven: What is it that we wish to achieve?
… It mentions image and HTML, we want SVG
… do we want to add wording that opens up other future possibilities?
Alain: output is about textnodes, not about children
Steven: You mean, if SVG is in an instance, could you use <output/> for that?
Erik: XML could be either serialised text, or as an instance
… how can you handle both cases with <output/>
… would an implementation be allowed to distinguish
… you have to decide how to distinguish.
… Presumably, <output ref="node" mediatype="..."/> that isn't base64, nor URI, but an SVG instance, you'd expect it to be used.
Steven: Split it three ways? URI, Base64, otherwise use the nodes?
Erik: If the nodes are compatible with the mediatype then use them
Alain: for instance mathml
ACTION: Steven to propose wording for the 3-way split for output @mediatype
AOB
Steven: I'm OK for next week, but the three after that I am at XML Summer School
[ADJOURN]