TD call time

<dsr> scribenick: dsr

<kaz> doodle poll results

Sebastian talks about the results of the doodle poll for the TD meeting slot

The poll results show that there isn’t a single slot that is good for all of those who responded.

Agenda

<inserted> TD agenda wiki

Welcome to the WoT WG's TD TF

Sebastian: I prepared some slides for newcomers.

Daniel: please send out a pointer to your introductory slides to the list for those people who couldn’t attend today

Sebastian: how widely? To the public IG list or the WG member list?

Dave: fine to send them to public list and to invite people to join the WG to help with the work

Sebastian shows us his slides

Motivation: thing description as device equivalent to index.html for web of pages

What kind of data do you serve? How can I access it? What protocol/encodings do you use? Who are you etc.

One server pushing a thing for another to consume

Thing description needs to be independent of platforms, programming languages, etc.

The publisher and consumer for a thing may be on different platforms and using different languages.

What is the core model we need to standardise? How to enable semantic interoperability? How to support data type definitions? Serialisation formats for thing descriptions. Support for security

We need a good framework for end to end security, e.g. to protect the thing description, communication channels, access control etc.

Introductions

We are all old hands, except for Achille

Achille: I am Achille Zappa from the Insight Centre in Galway, Ireland. We’re involved in other W3C groups

We’re interested in the Web of things and its potential, especially in respect to semantic modelling

He knows the rest of us so we skip a full round of introductions

Logistics

<inserted> TD wiki

Sebastian: asks about the wiki page, is this the same as for the IG, is there a Member only version?

<kaz> TD GitHub

Kaz: the above is the link to the Thing description github repository

Matthias suggested we use the regular mailing list with the [TD] prefix

So we don’t have a TD specific wiki, right.

Sebastian: happy with the TD GitHub

The member only mailing list is to be used for technical discussions

non-technical info could be shared on the public list

We will stick with the #wot-td IRC channel

I want to clarify the scribe before each call and announce this in advance

and I will rotate people for this in a fair way

We should use the GitHub issue tracker for our TD issues

<kaz> HTML rendered version of the TD draft

<inserted> scribenick: kaz

dape: relationship between this WG deliverable area and the old area?
... assume both the github areas are public

dsr: we should be careful to handle contributions from non-Members

kaz: right
... that's why we use ash-nazg for the GitHub repos
... there are already CONTRIBUTING.md and LICENSE.md

CONTRIBUTING.md

W3C Patent Policy FAQ 6

Sebastian: OK
... (and repeats the date/time for this TD call)

Outcome of the Santa Clara F2F Meeting

Santa Clara F2F Minutes

Sebastian: first discussion on the TD model

<dsr> I note that UML entity relationship diagrams such as with the tool preferred by Johannes, don’t deal with disjunctions which are definitely needed for a grammar for thing descriptions

Sebastian: grouping mechanism such as "#include"
... collaboration with IETF JSON Schema group (Henry Andrews)
... TD model is very important
... and would like to discuss that for next weeks

dsr: how we formally capture the requirements?
... actual documents? or something on the GitHub?

seba: requirements for the deliverables?

dsr: where the requirements for the spec came from
... e.g., due to OCF 1.1 spec we need something

seba: requirements could show up directly in the spec draft

dsr: it's usual to have use cases/requirements separately
... explain why we have which features

seba: requirements/use cases could be an issue on GitHub

dsr: I have another question
... diagram about UML entity relationship
... not complete yet
... what is the role of the diagrams?

seba: one day we need formal definition
... but most people would understand using diagrams
... the model can be powerful with formal definition, though
... diagrams are key tools
... and we should not require people special drawing mechanism
... Dave, you already have your idea for the model
... I also have some similar idea

dsr: we can invite Maria from the LD TF as well
... if you want
... we need to be careful about the contribution for the WG deliverable, though

seba: ok
... will be not available next week, so would like Taki to moderate the call

<dsr> Dave sends regrets for next week - I will be in Berlin for an EU project meeting

seba: Dave, can you contact Maria, etc.?

dsr: will contact them
... but regrets from myself

<achille_zappa> Achille sends regrets for next week - I will be in Berlin for an EU project meeting too

scripting notes

Nimura-san's GitHub issue suggesting we add use cases

kaz: regarding use cases/requirements, there was a similar discussion during the Script TF call today
... they will start with the GitHub issue above
... but may have a separate document for use cases/requirements
... might be an updated version of the existing use cases/requirements document

[ adjourned ]