W3C

– DRAFT –
WoT Use Cases

04 April 2023

Attendees

Present
Ege_Korkan, Kaz_Ashimura, Luca_Barbato, Michael_Lagally, Michael_McCool, Tomoaki_Mizushima
Regrets
-
Chair
Lagally
Scribe
McCool_

Meeting minutes

Housekeeping

McCool: last week we merged a PR on architecture, still have hazard use case on deck

Kaz: we should however discuss policy, but did not last week because Michael Lagally was not available

Lagally: appreciate it

McCool: think however we should "drain the queue", then close it while we discuss policy

Lagally: would prefer to defer and focus on policy

Lagally: let's look at the hazard PR at the end of the call

Policy

Lagally: let's review the current policy; we accept use case proposals widely, review as a group, then merge in to the document if appropriate
… have some templates (HTML and MD) with a small high-level requirements section
… then we did some editorial work to organize them, and in some cases asked contributors to consolidate and refine

McCool: think we need to define the overall process, and then the inputs, outputs, and contributors to each step
… for example, gathering use cases, extracting requirements, and prioritization and recommendations for work items

McCool: in summary, we focus on gathering inputs but not generating outputs

Ege: agree, we don't focus enough on outcomes and work items we should work on

Ege: focus of use case should be not to just gather iot use cases, but identify areas in which wot would be useful and have a business benefit

Lagally: agree, need phased approach, maybe different people involved

Kaz: most important part is identifying needs from industry
… although may have needs from different areas
… but should do requirements and gap analysis separately

Kaz: on the other hand, business benefit may be hard to define and may be too narrow

McCool: maybe we should consider "wide benefit", e.g. number of stakeholders impacted

Lagally: if we look retrospectively, see things that have been in the document for a while
… consider for example geolocation

Lagally: for example, we have known about geolocation for a while, but it was not dealt with

McCool: to be fair, we know about geolocation, but we have not done enough analysis of others

Ege: regarding geolocation, I don't really understand the requirements...

McCool: there are some documents on geolocation, but the analysis is not merged into the main document, so it's hard to find

Lagally: do we want requirements in a separate document or update this one?

McCool: personally think easier to have it all in one document, links etc. are easier

Kaz: agree requirements should be described separately, but there are relationships that need to be organized

<Mizushima> +1 kaz

McCool: current requirements lack links back to use cases

Lagally: these are horizontal and are quite old

McCool: perhaps we should have a new section for "use-case-driven requirements" and put the current set in a "broad" requirements section

McCool: also we should link the proposed work items back into use cases

McCool: and some of the work items were in fact driven by the coverage analysis

Ege: I did also analyse the coverage analysis but was only able to extract three actionable items
… one problem is lack of technical depth
… but for example, time series

McCool: so we need to do more work to extract technical requirements?

Ege: but in some cases we need to talk to the people who originally proposed a use case
… my opinion is that each use case must include *technical requirements*
… don't think current system is working well

Lagally: think it has been working well, have a good collection of use cases

Ege: personally, think success should be measured by successful specs

Lagally: but it has also been useful to gather interest

Kaz: for use cases as a starting point, should consider requirements as list of "pain points"
… what is difficult to do, we have to think about how to solve the problems to cover the requirements

McCool: personally I think use cases are about identifying problems clearly, but we still need to do work to identify solutions

Ege: don't think we should merge new use cases until we draw out the requirements

Lagally: don't however want a long queue of pending requests

Ege: but also bad to have a lot of use cases merged but now can't contact the people who submitted them
… and we also have a problem with redundancy and vagueness

Ege: when I read the use cases, a lot of them also are already satisfied

Lagally: you do have a concrete suggestion to hold off merging until we have requirements; still a little vague as a policy, since use case template does have a requirements section
… and who decides?

Ege: task force lead?

Lagally: consensus is better

Ege: also, spec writers and marketing have different needs

Lagally: think we need to think a bit more, consider the stakeholders, how the use cases will be moved, etc.
… but we need to improve the collaboration between groups

Kaz: agree consensus decision making better
… also, although we already have many use cases, we should also consider updates
… and we may want to consider restructuring

PR 211

Lagally: this is on Hazard Annotation

<luca_barbato> w3c/wot-usecases#211

<luca_barbato> https://github.com/w3c/wot-usecases/pull/211/files#diff-de6d415702ffd77dbeb76f19f38fb827983fef1113075d90e005b864b30f0a01R109

<luca_barbato> PR 211 - Hazard Annotation/files#diff-de6d415702ffd77dbeb76f19f38fb827983fef1113075d90e005b864b30f0a01R44

<kaz> (merged)

McCool: want to see if accessibility has been addressed; Yes, it has been addressed.

<kaz> [adjourned]

Minutes manually created (not a transcript), formatted by scribe.perl version 210 (Wed Jan 11 19:21:32 2023 UTC).