W3C

WoT Discovery

18 October 2021

Attendees

Present
Cristiano_Aguzzzi, Farshid_Tavakolizadeh, Kaz_Ashimura, Kunihiko_Toumura, Michael_McCool
Regrets
Christian_Glomb
Chair
McCool
Scribe
FarshidT

Meeting minutes

Minutes review

Sep 20th Minutes

<McCool> https://www.w3.org/2021/09/20-wot-discovery-minutes.html

no objections to publishing the minutes

Oct 4th Minutes

<kaz> https://www.w3.org/2021/10/04-wot-discovery-minutes.html

McCool: Farshid to be added to guests list for this meeting

no objections to publishing

PRs

PR: Clarify validation statements (#230)

https://github.com/w3c/wot-discovery/pull/230

McCool: It would be good to add an assertion in the validation section to clarify the directory does the validation. Also, to say that the directory MAY reject the request if the input is invalid.

https://github.com/w3c/wot-discovery/issues/99#issuecomment-945823118

no objections to merging

F2F action items

<kaz> Day 1

<kaz> Day 2

<kaz> Open Day 1

<kaz> Open Day 2

Simplified JSONPath

McCool: simple keyword search

McCool: could be added to appendix

Farshid: or non-normative section in place

Farshid: JSONPath is also used for selection

Farshid: keyword search could also be complicated (search in array, dictionary key/values, etc). Limiting to root level would make it simple

Cristiano: maybe simple keyword search (no jsonpath) would be better

McCool: Agrees to JSONPath keyword search, limited to the top level.

McCool: the keywork search would be a separate API e.g. /search/keyword?query=

McCool: maybe limited to title and descriptions

Farshid: string search is often not enough. What about datetime attributes?

McCool: we need to mandate at least one way to allow searching. The directory could contain millions of entires.

McCool: writing a issue to discuss the problem in detail

https://github.com/w3c/wot-discovery/issues/231

Farshid: there could be many different simple use cases. It is hard to cover all.

Kaz: we need to clarify the scenarios that we're addressing.

Farshid: full text search on millions of items could longer than querying all and doing efficient filtering on the client side

McCool: true but it would consume less networking resources

Next meeting

cancelled

Minutes manually created (not a transcript), formatted by scribe.perl version 147 (Thu Jun 24 22:21:39 2021 UTC).