W3C

- DRAFT -

MMI WG

07 Nov 2016

Agenda

See also: IRC log

Attendees

Present
Debbie, Dirk, Kaz
Regrets
helena
Chair
Debbie_Dahl
Scribe
ddahl

Contents


<scribe> scribe: ddahl

actions

action-454?

<trackbot> action-454 -- Deborah Dahl to Turn bullet list into text -- due 2016-09-05 -- OPEN

<trackbot> https://www.w3.org/2002/mmi/Group/track/actions/454

debbie: that was working on the use cases
... for robots

action-456?

<trackbot> action-456 -- Kazuyuki Ashimura to Generate a picture based on the tpac flip chart idea -- due 2016-10-10 -- OPEN

<trackbot> https://www.w3.org/2002/mmi/Group/track/actions/456

<kaz> Kaz's diagram

<kaz> SVG version diagram

<kaz> PNG version

debbie: what is the protocol between MMI and WoT?
... the line between the MMI and WoT boxes?

kaz: there are several possibilities

debbie: maybe Life Cycle events?

dirk: what is the main purpose of this?

kaz: MMI architecture could be the UI for the WoT framework
... EMMA could be used for interacting with devices
... WoT scripting also thinking about event handling and discovery, want to ask MMI to look at WoT proposal

debbie: we can look at that

<kaz> Intel's proposal for WoT scripts

kaz: a developer would be writing scripts to this API
... all the clients and servers need to use this API, not connect directly

dirk: arrow between the boxes needs to be a WoT protocol, not Life Cycle

kaz: this proposal doesn't handle events or data itself
... the detailed protocol is hidden

debbie: you could put a layer between the MMI and the WoT that would make the WoT look like a MC

kaz: the Automotive WG is looking at a low level socket interface rather than a scripting API
... this one is not directly handling protocols and data, but is more high level, it's more of a wrapper around a socket connection
... API definition depends on the policy of each standardization group
... W3C should clarify the layers of the socket interfaces
... low level, sockets, high level API's
... that should be handled by the TAG

debbie: is there a process for raising a question to the TAG?
... do we send an email?

kaz: many of the TAG visited Tokyo last week. I helped moderate a session about the Web Incubator Community Group

https://www.w3.org/community/wicg/

kaz: in the W3C we can send proposals to the TAG, but this is also a mechanism for the public

debbie: what's the question? what's the best way to evolve the web to non-traditional platforms

kaz: the web consists of browsers, but also servers and other related components
... browsers are important, but that's not all
... wondering how to integrate script discussion and UI discussion with MMI discussion

dirk: what is the problem between scripting and MMI?

kaz: the MMI architecture has a complete set of life cycle control, but the WoT group has to think about how to integrate devices and what kind of newly created script should be used
... how to integrate their proposals with our framework
... dirk, what kind of protocol do you use in your work, and how do you convert?

dirk: we used public/subscribe middleware so that we didn't care where something was deployed

kaz: what should we do if we need to access specific devices?

debbie: I've always used middleware between the IM and the device

kaz: MMI IM should be another WoT servient

dirk: is this more like an application using this servient?

kaz: WoT is an MC from the MMI viewpoint
... should there be one WoT servient that manages all the WoT servients?
... in one application?

debbie: if the WoT is one MC then we don't care about how it's organized internally

kaz: we need to know who to talk with

debbie: what are examples of those servients?

kaz: from the WoT viewpoint, each servient handles a capability and a device
...kaz: each servient should be attached to a device
... like a rice cooker or refrigerator
... the WoT box could contain the servients for the whole house

dirk: but we don't have the Russian doll kind of nesting like MMI

kaz: each module has server and client capability
... but some servients can act more like an IM and some are more like modality components

debbie: isn't the RM more of a processor?

kaz: RM could be split into processor and description

dirk: in MMI the behaviors are mostly defined by ExtensionNotification

kaz: we should respond to the WoT proposal and then we can have a joint discussion

debbie: is the scripting link a good place to start?
... someone needs an action to look the scripting proposal

kaz: we should look at the updated scripting proposal but we can start commenting on the discovery part

close action-456

<trackbot> Closed action-456.

<scribe> ACTION: kaz to update the WoT/MMI relationship diagram [recorded in http://www.w3.org/2016/11/07-multimodal-minutes.html#action01]

<trackbot> Created ACTION-460 - Update the wot/mmi relationship diagram [on Kazuyuki Ashimura - due 2016-11-14].

action-457?

<trackbot> action-457 -- Kazuyuki Ashimura to Make a strawman/template diagram for relationships between entities in a use case -- due 2016-10-10 -- OPEN

<trackbot> https://www.w3.org/2002/mmi/Group/track/actions/457

kaz: will do that
... also need to merge dirk's update to the use cases

action-459?

<trackbot> action-459 -- Kazuyuki Ashimura to Coordinate with helena about her library software for the github -- due 2016-10-31 -- OPEN

<trackbot> https://www.w3.org/2002/mmi/Group/track/actions/459

kaz: will coordinate with helena
... helena could create a subfolder of the MMI repository, but I need to talk with her about licensing

<kaz> Meeting: MMI WG

Summary of Action Items

[NEW] ACTION: kaz to update the WoT/MMI relationship diagram [recorded in http://www.w3.org/2016/11/07-multimodal-minutes.html#action01]
 

Summary of Resolutions

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.148 (CVS log)
$Date: 2016/11/07 15:15:13 $