14:57:39 <RRSAgent> RRSAgent has joined #voiceinteraction
14:57:39 <RRSAgent> logging to https://www.w3.org/2022/06/01-voiceinteraction-irc
14:57:52 <ddahl> meeting: Voice Interaction Community Group
14:58:00 <ddahl> chair: debbie
14:58:09 <ddahl> scribe: ddahl
15:10:24 <ddahl> present: debbie,michael,noreen,bev
15:10:42 <ddahl> topic: promotional work
15:11:10 <ddahl> debbie: jim put together a proposal for voice workshop at W3C
15:11:27 <ddahl> ...should try another way to promote that
15:13:15 <ddahl> michael: jon was at the European Voice Lunch
15:15:26 <ddahl> debbie: check to see if I sent bev the 100 word abstract
15:17:09 <ddahl> debbie: can resend
15:17:15 <ddahl> ...if you can use it
15:18:23 <ddahl> debbie: the second topic was reviewing the interfaces document
15:19:18 <ddahl> ...but let's skip that and talk about the MMI Architecture/Interfaces comparison
15:26:05 <ddahl> debbie: reviews MMI architecture
15:27:17 <ddahl> ...sessionid and Context seem to be the same
15:28:10 <ddahl> ... requestid and RequestID are the same
15:29:05 <ddahl> ...MMI has a Data event vs. three fields for Interfaces document
15:34:49 <ddahl> ...Content/ContentURL isn't in the Interfaces document
15:36:22 <ddahl> jim: we shouldn't tell the component what to do
15:38:17 <ddahl> debbie: maybe we used to have a single system communicating vs now we have more collaborating systems
15:39:21 <ddahl> jim: only one message type (Start) vs. collection of messages where the name implies the action
15:43:41 <ddahl> jim: only one message type, with the first field the action that's being requested and the other fields are parameters for that action
15:45:00 <ddahl> debbie: one messge e.g. "voiceInteractionMessage", with an action like start, stop, acknowledge
15:46:25 <ddahl> debbie: fields could be contingent on actions
15:46:55 <ddahl> jim: a specific message type has a fixed set of fields
15:48:26 <ddahl> ...could we define new messages?
15:52:21 <ddahl> noreen: start is like a presence detection message in a chatbot
15:52:54 <ddahl> debbie: we should think about presence detection
15:54:04 <ddahl> jim: how do these instructions map to messages?
15:56:04 <ddahl> debbie: how generic do we want the messages to be?
15:56:24 <ddahl> jim: the messages should be pretty generic, like in MMI Architecture
15:57:08 <ddahl> bev: messages should be kept at a high level
15:57:35 <ddahl> ...do we have a list of protocols?
15:57:49 <ddahl> debbie: like HTTP or REST?
15:58:30 <ddahl> ...try to stay at top of protocol stack
15:59:54 <ddahl> debbie: we should list messages
16:00:21 <ddahl> jim: we should call these actions
16:00:27 <ddahl> debbie: what about events?
16:02:31 <ddahl> bev: thinks actions is good
16:03:23 <ddahl> jim: could be messages, actions, or events
16:04:56 <ddahl> rrsagent, format minutes
16:04:56 <RRSAgent> I have made the request to generate https://www.w3.org/2022/06/01-voiceinteraction-minutes.html ddahl
16:05:01 <ddahl> rrsagent, make logs public
17:14:06 <ddahl> ddahl has left #voiceinteraction