<kaz> scribenick: dape

Agenda

<kaz> McCool: (checks the agenda for today)

Review of PRs

McCool: Was not able to finish my PR about report generation

<kaz> PRs

<inserted> PR 16

McCool: let's merge validation PR#16
... validation scripts are now in
... report generation needs more work

<kaz> PR 28

Farshid: worked on tool for TD sync
... python script does the work

McCool: Suggest merging PR
... tool re-ordering can be done later if we have multiple tools
... We also need to document SoftEther setup
... let's create issue for that

Ege: Pushed Wade results already... without PR
... last column in WADE.csv is the result

<kaz> WADE.csv

McCool: Can we organize it based on manual assertions
... we should have kind of these results in general.. not only about WADE

Ege: We can restructure the presentation

McCool: We should have a way to create a test dashboard automatically... with categories et cetera
... rigth now we have single pass/failed. we might want to expand the information.

Ege: Could think of "op" field keyword, protocol, security

McCool: Suggest expanding template. Add TODOs in data/interop/README.md

Ege: For complete testing we can still use test-bench

McCool: CSV should be usable for manual editing but may also contain additional information
... DP added id's for Siemens TDs

Daniel: Correct, to overcome auto generated id creation issue by directory

<inserted> Issue 22

McCool: related to issue#22

Wrap-up

McCool: on Tuesday during VF2F we have PlugFest report
... please add comments to issue tracker
... e.g., yesterday we discussed xpath vs. Json path
... see issue #29

<inserted> Issue 29

Ege: XPath supports JSON querying .. in recent version
... issue exists in discovery repo

Farshid: New version supports arrays

McCool: JSON path might be less powerful ..

<Ege> https://github.com/w3c/wot-discovery/issues/17#issuecomment-621190733

Farshid: Yes, but there is no formal specification either...
... there is also difference in implementation supporting certain features

<kaz> JSONata

MR: There is also JSONata, used by nodered

McCool: Miguel, please comment in issue

MR: will do

McCool: 2 outcomes so far
... 1. primary key
... 2. XPath features

Farshid: Issue with types, ..thing vs. directory
... will create issue

<miguelrk> JSONata: an alternative JSON query and transformation language (https://jsonata.org/)

McCool: Need to figure out how many types we need
... information model needs to contain timestamp and id
... let's discuss this topic in discovery
... One other topic is VPN. Question: was it useful?
... VPN was essential for :
... - using mDNS hostnames to avoid issues with IP changing
... - using local DNS-SD for discovery
... Future testing:
... - mapping from fixed URL
... Useful:
... - testing NAT traversal

Ege: Not sure who was dependent on VPN?

McCool: easier to bring up device inside VPN
... less security troubles
... no security on the Internet would not be a good idea

Ege: Device owner issue

McCool: For PF report we do not have much time

Sebatian: w.r.t. to XPath, JSON path. Did you look into JSON LD prefixes and the related issue

McCool: looked at it in discovery TF
... we have various issues here

Issues identified during the online PlugFest

Sebatian: Can discuss with JSON-LD group

Koster: Need light weight query solution

McCool: We need standard, JSON path isn't

Koster: there is also JSON Patch

JSON Patch

McCool/Koster: We should start with uses case

McCool: Any PF outcome for report

<Zakim> kaz, you wanted to suggest we add "PlugFest outcome" label as well to issue 29 and 22 (and some more possible ones)

Kaz: suggest adding label to issues w.r.t. PF outcome

McCool: Good idea

Farshid: I also have some slides. I will use them on Monday

Kaz: Do we want to invite JSON-LD people?

McCool: packed.. should definetly invite them for discovery call
... DID is coming to VF2F

Koster: We could let JSON-LD know that we have F2F... maybe one has time to show up

McCool: One other issue is having validation/report ready before PF

Koster: Are we planning to go on with validation... with experimental stuff

McCool: We also had prototypes folder

Koster: Makes sense

McCool: Kaz, can we keep VPN up?

Kaz: checking with the system team

Koster: I also think an ongoing activity makes sense

McCool: Agree
... having an ongoing process makes sense
... devices up would be good
... might need some money to keep server up. Can pay for it
... at least we should archive image