IDE Minutes 2009 05-12

From MemberWiki

Jump to: navigation, search




  • Jon Ferraiolo, IBM
  • Rich Thompson, IBM
  • Lori Hylan-Cho, Aptana
  • Howard Weingram, TIBCO
  • Kin Blas, Adobe
  • Javier Pedemonte, IBM
  • Adam Peller, IBM


Topic: Rich's email on modes

Jon: Rich simply suggests adding some clarification words to the definition of edit mode because IBM has found the need for multiple phases of edit modes.

Rich: Yes. In our internal spec, we'll add a couple of words to be more precise about what edit mode means versus other modes

Jon: Can you forward your words?

Rich: Yes

Jon: The 3 levels of edit mode make sense to me.

RESOLUTION: Add colored note to spec about including more precise wording for edit mode

ACTION: Rich to forward the wording that IBM is using

Topic: Intraspec links

Jon: After talking about this for a long time, I finally created a wiki extension that allows intra-spec hyperlinks. Couldn't just use normal hyperlinks because the root name of the wiki page might change as we go from version to version.

Lori: Hooray!

Jon: I started using this extension in a couple of places, but still need to change lots of things. Will help simplify the spec.

Lori: And leverage the Web.

Topic: Widget Overview chapter

Jon: I did a bunch of spec cleanup. As we have already discussed, we needed a widget primer of sorts. I have started down that path by creating an overview chapter. I took the intro sections from the Widget Metadata chapter and moved them to the overview chapter. Then I cleaned up various things. Need a shorter example. The current one was copied out of Dojo but is too long.

Jon: I also created an overview chapter for the JS API features. That overview chapter isn't as far along.

Jon: The new structure of the spec: (a) Intro chapter, (b) Widget chapters (overview, metadata, APIs), (c) API Metadata chapters (overview, metadata), (d) shared chapters.

Jon: I consolidated the Topics chapter into the Widget Metadata chapter since only widgets use Topics. Any objections?

(no objections)

Jon: The rest of the chapters contain markup that is shared by both the widgets and the API metadata. But wait, not sure about the Compatibility chapter. Maybe that's only used by API metadata. If it is only used by API metadata, then OK if I remove that chapter and move the old contents to the API Metadata chapter?

(no objections)

Topic: Tables for attribute definitions

Howard: It is really difficult to read the attribute definitions. Would it be possible to put attribute definitions into HTML tables? I'm not sure we have provided default values for all of the attribute values. The definitions may not always be complete.

Jon: I like the table idea. Can't believe it didn't occur to me earlier. Let me play with it and see how it looks.

Lori: I always find it easier to edit tables in an HTML editor, then paste it in.

Topic: JSDoc Documentation conventions into appendix

Jon: We decided to do this yesterday in the Interop phone call about the Hub. Same thing here?

Howard: Yes

RESOLUTION: Move JSDoc conventions out of the Widget API chapter

Topic: publish/subscribe attributes

Jon/Rich: IBM discovered that these weren't needed in most scenarios and just made things complicated. Best to remove these attributes.

Howard: I need more time. No deep philosophical objection, but want to think about this some more. May want a readonly facility.

Jon: We do have a 'readonly' attribute on the property element

Howard: I'm thinking of something else.

Topic: getId() versus getWidgetId()

Howard: Need to somehow namespace the API name in case of mixins

Adam/Jon: But this API is already on an OpenAjax subobject

Howard: OK, then getId() is fine

RESOLUTION: API name is getId()

Topic: BIDI

Jon: Are we done now in getting feedback from IBM's Middle East people?

Adam: I think so.

Jon: Good news is that they looked at the OpenSocial BIDI features in detail and felt that the OpenSocial approach was good. It's just that they were agonizing about whether to extend the OpenSocial approach, particularly for contextual directionality.

Adam: Also really wanted to deal with shaping, but they realized that one is too hard to address today given the browser world.

Jon: Adam and I concluded that you don't have to use the OpenSocial BIDI syntax and can do whatever you want to handle BIDI within your own JavaScript. We can extend the markup features in the future and require in v1 that if those features aren't sufficient, then take care of things in JavaScript

Howard: Yes. Push it off to a future release based on demand from the industry.

Rich: We need to be careful about what we require from implementers.

RESOLUTION: Adopt OpenSocial BIDI features wholesale, with no additions

Topic: Restoring the "reference implementation"

Jon: Just an announcement. Last fall, we posted a open source mashup application that we called the "reference implementation" that supported OpenAjax Widgets and OpenAjax Hub. That implementation is now out of date with the latest specs. Javier is updating refimpl now to match Hub 2.0 and the latest Widgets spec. It will include the open source widget loader that Adam has developed.

Personal tools