From MemberWiki

Attendees

• Stew Nickolas, IBM
• Lori Hylan-Cho, Adobe
• Jon Ferraiolo, IBM
• Ted Thibodeau, OpenLink
• Ingo Muschenetz, Aptana
• Kevin Hakman, Aptana
• Bertrand Le Roy, Microsoft
• Phil Berkland, IBM
• Rich Thompson, IBM

Minutes

Kevin: Last week made good progress. Got about 10 or less items to discuss. Some open items--Betrand to draft a proposal on those.

Bertand: I unfortunately didn't have time

Kevin: Moving onto <remarks>

Ingo: It's basically an extended description for more information

Kevin: Laurie and Bertrand, is that something you find necessary?

Bertrand: We have that element

Lori: Where is that useful? I can think of a use case where we could use it, but we're not using it right now

Bertrand: It's one of those things that allows you to structure your description more than plain text

Jon: What are the requirements on the IDE?

Kevin: It's optional for the tool to display and in whatever format the tool decides

Lori: My only quibble was with the name remarks. It seems very informal

Jon: What name would you choose?

Lori: Unsure

Ingo: We chose it because it was similar to other items, such as in JavaDoc

Bertrand: Is it a sibling or child of description?

Ingo: In our case, it's a sibling

Jon: I'd suggest sibling pending Betrand's research on what Microsoft does. What elements does it attach to?

Ingo: We use it on class, property and method

Kevin: return-description: So that becomes a child of the <returns> element. Any objctions to that?

Kevin: next item is @scope

Ingo: Is that a hinting mechanism for the documentation generation?

Jon: it seems to be useful in the JSDoc to XML generation.

Kevin: It seems redundant to our current scope attribute. Any objections to removing it? Okay, removed.

Kevin: @sdoc. This seems like an Aptana thing.

Ingo: It's to reference an external documentation file

Kevin: @see: Waiting on a proposal from Bertrand

Bertrand: I've asked, but not had a response yet

Kevin: @since. Seems like an Aptana thing

Jon: Looking before the meeting, it's seems like we may made a mistake in omitting availablility. It could just be text

Kevin: I agree it might be useful to just capture the text.

Kevin: Any objection to reinstating availability and since?

Rich: Perhaps 'available' is a better term?

Jon: To confirm, 'deprecated' and 'available' elements with version attributes and text descriptions? So the specifications node is an Aptana specific thing?

Ingo: Yes, I think we can do that using extensions

Kevin: Next item in the matrix is static? I think we've handed that already

Ingo: It's the same, useful for JSDoc to HTML generation

Kevin: Moving on to value and values. It's effectively enumerations

Bertrand: In Microsoft, the enumerations are just fields. This "values" construct works but requires a bunch of repetition. We're fine with it, but one concern is how would other IDEs be able to look at the enumeration type and show the enumerations?

Kevin: Doesn't a concept of a class capture that?

Bertrand: Not exactly.

Ingo: We don't have that much repetition

Bertrand: We make great use of enumeration types. There are also enumerations and flags

Kevin: Is this an area where you'd like to put together a proposal? Something where there are both of those cases would be useful to show. Thanks bertand

Jon: Can I make a draft consensus about <values><value>?

Ingo: We need to make an example of using flag or flags as Betrand had suggested

Kevin: Jon, for now can we make some semantics about what to call.

Kevin: @version. Looks like an Aptana thing

Ingo: it's the version of the whole library, i.e. jQuery 1.2.3

Kevin: Any objection to including that? This would be an attribute on the API element. Okay, approved.

Kevin: Items left: "type" attribute on <alias>

Ingo: We used that in a very specialized circumstance which isn't required here. It can be removed.

Kevin: How about the <browsers> element?

Ingo: I think we changed it to the useragent element.

Bertrand: What Aptana has done here is well-thought out, and will work well here. How does the versions attribute work?

...some discussion about possibilities about the version attribute... end result is "-3.0" is less than, "3.0" is exact, "2.0-3.0" is range, "3.0+" is equal to and greater than

Kevin: We're at the end of the hour. We'll pick up discussion next week.