Meeting minutes
Previous Actions
Action: Steven to research where to put S for attributes.
Action: Michael to comment on conformance section of new draft
<trackbot> Sorry, but no Tracker is associated with this channel.
<trackbot> Sorry, but no Tracker is associated with this channel.
Steven: My action continues
Steven: Michael's is done
… on the agenda
New draft
https://
https://
https://
Steven: Do want to discuss?
… or should I just add to spec
https://
https://
https://
MIchael: As questions arise please ask.
Action: Steven to apply Michael's text to spec
<trackbot> Sorry, but no Tracker is associated with this channel.
Michael: Are people happy with the suggestion of giving a choice about which tree gets returned for ambiguity.
… I want to do that.
… I would rather it be conformant
Steven: Processors may on user option return more than one
[General agreement]
John: How about distinguishing when there are several?
Steven: Already there.
… attribute
Tomos: Can we make that optional?
Michael: it is only an attribute
Steven: In the ixml namespace
Tomos: Can we make it optionally switchable?
Michael: I am uneasy, but if it is a user option...
Tomos: It isn't required, but implementers can decide to make it option
Michael: I think I can live with that.
Action: Steven to find wording to make ambiguous attribute optional
<trackbot> Sorry, but no Tracker is associated with this channel.
Namespaces
https://
Tomos: I want to support namespaces
… take the roundtripping.
… to take XML and turn it into another form.
Michael: mathml to latex and back?
Tomos: If you can express it with a grammar then the two abstractions are equivalent.
Steven: It sounds like we are redesigning ixml
… adding new requirements
Tomos: No, I want to plug some gaps.
Steven: So you would be happy if we could insert strings into the serialisation
Tomos: Just namespaces would be enoughj
John: a different example. Can you parse XML?
Steven: Yes.
John: Then you can handle namespaces.
Michael: Let me try and understand the question.
S/Then/Then/
Tomos: I would like to accept prefixed names
Steven: They already are aren't they?
Tomos: Even if the prefix is available, we can't specify the namespace
… that's all we need
… I don't think it's a big change
Michael: This is going to be a judgement call of cost:benefit, and if it complicates things. ixml has relative simplicity. If it is low-cost, there are advantages for xml. We need to have proposals on how to do it.
Michael: I would rather make a decision once we had a proposal or two
Tomos: Can we post on the github the grammar
Steven: Isn't it there already?
Tomos: I mean a canonical file.
Action: Steven post grammar on github
<trackbot> Sorry, but no Tracker is associated with this channel.
Michael: I propose that people take actions to sketch out a proposal
Action: All to propose draft changes for namespace support
<trackbot> Sorry, but no Tracker is associated with this channel.
Requirements:
https://
Steven: "Should we explicitely state the requirements for ixml?"
Tomos: Yes.
John: Roundtripping requires you keep all the information in the result.
Steven: Grammar author's responsibility.
<cmsmcq> Is round-tripping a requirement? Is round-trippability a requirement? John - but when the grammar is used as a kind of query to extract important bits, we often throw information away.
John: A double round trip would be idempotent
MIchael: I'm unsure about formalising the requirements. The discussions are not always useful.
Steven: Requirements help guide discussions.
Action: Steven draft a list of the requirements as they were
<trackbot> Sorry, but no Tracker is associated with this channel.
Tomos: I'm in favour of controlling how the XML is produced.
Michael: Is the requirement "Control over what XML is produced" or "Moderate control"?
AOB
Tomos: I will be happy to take the minutes next time.
Steven: I will be backup
Steven: Next meeting Tues 8 June, 14Z
rssagent, make minutes