14:58:33 <RRSAgent> RRSAgent has joined #wot-discovery
14:58:33 <RRSAgent> logging to https://www.w3.org/2021/11/15-wot-discovery-irc
14:58:41 <kaz> meeting: WoT Discovery
14:59:30 <kaz> present+ Kaz_Ashimura
14:59:42 <glomb> glomb has joined #wot-discovery
15:00:09 <kaz> present+ Christian_Glomb
15:00:11 <acimmino> acimmino has joined #wot-discovery
15:01:18 <kaz> present+ Andrea_Cimmino
15:01:33 <kaz> present+ Cristiano_Aguzzi
15:01:47 <cris> cris has joined #wot-discovery
15:02:46 <kaz> present+ Michael_McCool, Farshid_Tavakolizadeh, Kunihiko_Toumura
15:02:53 <ktoumura> ktoumura has joined #wot-discovery
15:03:00 <kaz> rrsagent, make log public
15:03:02 <Mizushima> Mizushima has joined #wot-discovery
15:03:04 <kaz> rrsagent, draft minutes
15:03:04 <RRSAgent> I have made the request to generate https://www.w3.org/2021/11/15-wot-discovery-minutes.html kaz
15:04:29 <kaz> Agenda: https://www.w3.org/WoT/IG/wiki/WG_WoT_Discovery_WebConf#15_November_2021
15:04:32 <McCool> McCool has joined #wot-discovery
15:04:58 <kaz> present+ Tomoaki_Mizushima
15:05:16 <kaz> scribenick: glomb
15:05:29 <glomb> https://www.w3.org/2021/11/08-wot-discovery-minutes.html
15:05:42 <kaz> i/minutes/topic: Minutes/
15:09:00 <glomb> Approved
15:10:26 <glomb> Agenda approved
15:10:30 <glomb> mccool - Update on JSON path
15:11:01 <glomb> IETF- still draft
15:11:15 <glomb> Important issues:
15:12:09 <kaz> i/Agenda approved/topic: Agenda/
15:12:18 <kaz> i/Update/topic: JSON Path/
15:13:27 <glomb> - Selectors applied in sequence; each has input and output; when searching for keywords - no regex selector yet defined - exact comparison required
15:14:03 <glomb> - "in": Can look for number or path but not string in string
15:15:02 <glomb> AI for us: generate some "typical" queries needed
15:15:24 <glomb> Cristiano agrees on found issues
15:16:08 <kaz> i|Update on|-> https://www.ietf.org/archive/id/draft-ietf-jsonpath-base-02.txt JSONPath - draft-ietf-jsonpath-base-02
15:16:16 <kaz> rrsagent, make log public
15:16:22 <kaz> rrsagent, draft minutes
15:16:22 <RRSAgent> I have made the request to generate https://www.w3.org/2021/11/15-wot-discovery-minutes.html kaz
15:16:29 <glomb> Cristiano - should we continue w/ own query mechanism?
15:17:02 <kaz> chair: McCool
15:17:07 <glomb> mccool: Several options: - Only rely on Xpath, but only one implementation at the moment
15:17:19 <kaz> s/Cristiano -/Cristiano:/
15:17:38 <glomb> - Or: define own query mechanism
15:18:31 <kaz> i/topic: JSON Path/no changes to the agenda/
15:18:31 <glomb> Cristiano: Yes, but should be simple to implement, just filtering instead of search
15:18:34 <kaz> rrsagent, draft minutes
15:18:34 <RRSAgent> I have made the request to generate https://www.w3.org/2021/11/15-wot-discovery-minutes.html kaz
15:19:14 <glomb> See: https://github.com/w3c/wot-discovery/issues/231
15:19:15 <kaz> s/mccool -/mccool:/
15:19:20 <kaz> i/no changes/scribenick: kaz/
15:19:36 <kaz> i/Update on/scribenick: glomb/
15:19:39 <kaz> rrsagent, draft minutes
15:19:39 <RRSAgent> I have made the request to generate https://www.w3.org/2021/11/15-wot-discovery-minutes.html kaz
15:20:24 <glomb> mccool - Need to support substring search in title and description
15:20:47 <kaz> s/mccool -/mccool:/
15:21:11 <glomb> Cristiano: regex also not (well) supported in JSONpath implementation, so might be "at risk"
15:21:42 <kaz> q?
15:22:17 <glomb> Farshid: Move JSONpath as informative and then if it is an RFC move to normative
15:24:50 <glomb> See summary in: https://github.com/w3c/wot-discovery/issues/231
15:25:42 <cris> https://cburgmer.github.io/json-path-comparison/#filter_expression_with_regular_expression
15:26:00 <FarshidT> FarshidT has joined #wot-discovery
15:28:38 <Mizushima_2> Mizushima_2 has joined #wot-discovery
15:29:47 <McCool> example of what I'd like to see: ?("light" in @.description) as a special-purpose substring operator
15:33:05 <glomb> Farshid: What about exact key-value matching? (JSONpath "equals" operator)
15:33:27 <FarshidT> examples: https://github.com/linksmart/thing-directory/wiki/Query-Language
15:33:32 <McCool> or we could just do exact: ?("Light" in @."@type")
15:36:41 <glomb> Could return only ID or array of IDs and then get the TD(s) based on the IDs
15:36:51 <kaz> q?
15:37:16 <kaz> q+ acimmino
15:37:17 <kaz> ack a
15:37:21 <kaz> q+ glomb
15:37:23 <kaz> ack g
15:37:23 <cris> q+
15:38:12 <glomb> Andrea: Getting ID always sufficient?
15:38:57 <kaz> q?
15:39:05 <glomb> E.g. for discovering endpoints - then not complete TD needed
15:39:21 <kaz> ack c
15:40:18 <glomb> Cristiano: let's define a really small subset of operation - for more complex queries use SPARQL
15:42:15 <kaz> q?
15:43:04 <glomb> Farshid: We should not create a new query language, rather use (JSONpath) subset and then go to full JSONpath once it is an RFC
15:44:13 <glomb> E.g. exact string matching
15:46:20 <glomb> Do JSONpath operators also work with dates?
15:47:15 <glomb> But JSON itself has no "time" type
15:48:08 <glomb> Epoch time comparison would work
15:49:49 <glomb> Cristiano already prepared a proposal for option 2
15:56:19 <glomb> mccool: Option 2 has a dependency on JSONpath spec - even if we pick a limited subset
15:57:16 <kaz> q+
15:58:08 <glomb> Possible plan: Drop JSONpath and have Xpath and SPARQL as optional
15:58:53 <glomb> Farshid: Could keep JSONpath as non-normative because it is already there
15:59:00 <kaz> q?
15:59:41 <glomb> If non-normative doesn't need to be tested - will just be moved in the spec.
16:01:35 <kaz> kaz: also please clarify the section structure and think about which sections would be normative at least by the end of Nov (as agreed)
16:01:39 <kaz> q?
16:01:43 <kaz> ack k
16:01:54 <glomb> Andrea: where to put examples - will volunteer for SPARQL
16:02:32 <glomb> PR about this discussed next time
20:56:01 <Zakim> Zakim has left #wot-discovery