From MemberWiki

URL: http://www.openajax.org/member/wiki/Minutes-Special-DeviceAPIs-20080506

Contents 1 Attendees 2 Agenda 3 Minutes 3.1 Action list status and whether we can move forward 3.2 OMTP coordination 3.3 Shim layer 3.4 Wrapup

Attendees

• Paddy Byers, Aplix
• Guillermo Caudevilla, Vodafone
• David Pollington, Vodafone
• Jon Ferraiolo, IBM
• Bob Goodman, IBM
• Krishna Shankar, Cisco
• Kai Hendry, Aplix
• Arun Ranganathan, Mozilla

Agenda

• Agenda
  • Review action list - where do we stand?
    • Mobile Device APIs initiative todo list
  • Is there anything else we need to do before claiming victory on our Exploratory phase?
  • Proposed way forward:
    1. Publicize our progress. I don't think we can justify/execute a press release on this particular topic. Instead, send short announcement articles to Ajaxian and AJAXWorld, and also author/submit a 2000-world article to AJAXWorld magazine (for both print and Web).
    2. Start work on a Working Group charter that describes what formal activities we should take up at OpenAjax Alliance. Here is my thinking of what the charter might say:
       • Work on a lengthy security white paper about allowing the Web Runtime to access device apis
         • But need to factor in coordination with OMTP
       • Launch an open source project for a JavaScript shim layer
         • Need to sketch out a high-level plan (who, what, when, how)
       • Document the shim layer within a Specification
    3. Send email to selected major players to tell them we finished the exploratory phase, please review and send feedback
  • Decide on next meeting and frequency of future meetings
  • Full list of our Mobile Device API wiki pages:
    • Mobile Device APIs initiative home page
    • Mobile Device APIs Objectives
    • Mobile Device APIs Use Cases
    • Mobile Device APIs Requirements
    • Mobile Device APIs Security
    • Mobile Device APIs Due Diligence Industry Survey
    • Mobile Device APIs initiative todo list

Minutes

Action list status and whether we can move forward

(Jon talks about how all actions have been completed except maybe one, which is where Nikunj volunteered to look at requirements and see if there are any use cases that need to be added, which Jon said was a nice-to-have action. Jon asks if we can consider the exploratory phase complete and move forward to the next phase)

Krishna: I think we can move forward.

(no other comments)

Jon: OK, let's talk about how we might move forward.

OMTP coordination

Jon: But first let me report on a coordination phone call I had just an hour ago with Nick Allott and Caroline Belrose regarding OMTP's work in the same general area of device apis. They are working on a similar white paper as we are thinking about in the same sort of timeframe.

Paddy: There was an OMTP f2f meeting this week where we talked about two strands of work. Architecture and Security, and then interfaces. My opinion is that in terms of maturity of thinking, very similar to here. Not surprising as there are overlapping contributors. The two organizations have somewhat different industry representation. They have mainly operators and device manufacturers. They are short on browser representation.

Jon: They are also working on device APIs, but there is likely to be a difference there because we are thinking about a JavaScript shim with adapters, whereas they are more likely to specify standard APIs that device manufacturers should implement.

Paddy: Yes, they aren't working on the Ajax layer.

Paddy: Historically, operators make requests of device manufacturers about what should be embedded on devices. But emphasis now is to prevent duplication and make sure there are no gaps, but still feeling their way.

Jon: Makes sense.

Krishna: How about a joint paper with OMTP?

Jon: Yes and no. Yes, we can definitely collaborate. But we probably can't actually do joint publishing from both organizations simultaneously, if only because of IP policy differences

Paddy: By end of year OMTP is shooting for either a precise spec or a real reference implementations

Shankar: A real product implementation?

Paddy: Can't say at this time. But the spec would be capable of being implementable and will try to get implementations. Other question is browser vendors who are not in OMTP.

Jon: Nokia can be a proxy for WebKit. Is Mozilla in OMTP?

Arun: Not to my knowledge.

Jon: Opera?

Paddy: Not to my knowledge

(no discussion about MS)

Jon: Paddy, you are in both. What do you think we should do on our end?

Paddy: I have contributed the same things to both. Caroline is chairing the OMTP effort. So far, the recommendation to OMTP will be aligned with what we have at OpenAjax. But there may be differences ultimately in scope. OMTP has usually given specific security approaches.

Jon: Yes, now the vision is aligned, but things might change in the future. How about letting OMTP take the lead for now and we help, and then wait and see how things go?

Krishna: That would allow us to focus on other things. If overlap is 50-60%, then yes that would make sense.

Paddy: I would be cautious about ceasing activity. OMTP has important constituency, the carriers, but equally there is negligible input from browser manufacturers. And I worry about them being open enough.

Jon: Wait and see?

Krishna: Might lose some momentum that we now have. Could maybe do a white paper in a couple of months?

Jon: Last year we did a major white paper in less than 2 months, so it is possible, but not guaranteed. The mobile white paper has been going on 9 months, is done, but still waiting for me to post it.

Paddy: Who is prepared to make the effort? If me, might as well happen in OMTP.

Paddy: We could get a liaison statement with OMTP rather than do work here

Jon: Nick is trying to get OMTP to join OpenAjax Alliance to help with liaison effort. My thinking is wait and see and then look at what gaps start to emerge

Krishna: Yes. Look again in a month or so.

Paddy: Yes, that's fair.

Shim layer

Jon: Now let's talk about the shim layer.

Arun: Are people aware of LocationAware from Mozilla? We would like to standardize on ECMA APIs for location

Kai: Google Gears location APIs are in the WAF WG

(someone): WAF? Not WebAPIs instead of WhatWG?

Jon: Shim layer would require lots of contributors. More difficult than a normal Ajax effort because most of the work is in the adapters.

Kai: I would be willing to contribute with Aplix WebVM. Maybe I would work with Opera. I'm also kind of interested in Gears.

Jon: IBM might contribute, but no approvals at this time. How about Vodafone?

Guillermo: Would have to talk but maybe yes with MobileScript source code.

Jon: For Windows Mobile and Symbian?

Guillermo: Yes, have things working on Windows Mobile and Symbian. One version with SpiderMonkey. Also working on a WebKit version.

Jon: We need to make sure we have critical mass.

Krishna: Maybe not when we start. People might join once they see things going.

Jon: Yes

(Kai suggestions about Silverlight and Flex involvement)

Jon: Yes, but priority is with Open Web vendors.

Arun: We have the Fennec project. Want to explore device APIs. Our vision is web is one platform. Interested in location, also address book. Mozilla now runs on a high-end Nokia device. The LocationAware effort is by the same guy who did Minimo.

Kai: We would like to see Mozilla have a plugin API that can be used on an experimental basis.

Wrapup

Jon: OK, out of time. So, we are moving onto the next phase. Need to have a plan for who/what/when/how for the shim layer. I'll start something on the wiki. I'll ask the chairs for when we will have our next meeting. Maybe next week, maybe the week after.