– DRAFT –
WoT Discovery

review minutes

meeting starts reviewing the minutes

<kaz> Sep-26

review PRs

PR #422

<kaz> PR 422 - Clean up Siemens/Logilab results

this PR was merged, it updates the siemens logilab implementation results

<kaz> Rendered version of the Draft Implementation Report

McCool: in general cr transition will be difficult to be done with current implementation results, many features are at risk

McCool: it seems that only one implementation provides SPARQL capabilities. Nevertheless, there should be two.

McCool: it seems that wothive implementation used ";" instead of "," for the results and that messes up the table

mccool fixes the file

McCool then updated the table, now it seems more of the SPARQL features are covered

the new table is rendered

McCool: upd still has a lot of features uncovered

Toumura: could you check the request I made in wot-test ? #443

<kaz> wot-testing PR 443 - Update Implementation Description etc.

PR is merged adding the assertions

mccool updates the table with the new implementation report

McCool: it seems that no at risk is changed with this new report

Cristiano: there are two problems. One is the fact we do not check the code 404 for the SPARQL

mccool adds the ttd-unsopported-feature to the manual assertion csv

Cristiano: the second is about the anonymous id

McCool: it seems that at the time is been tested no implementation performs this operation

Cristiano: I think farshid's implementation should provide this feature

also, cris implementation should cover this feature

McCool: the spec should be clearer about this

McCool: cris and farshid take over this

McCool: now we have the manual test
… only two implementations are describe as discoverer
… we may need more
… I'm concern that node-wot does not implements discoverer

Cristiano: we are currently working on this, probably it will be done in two weeks

McCool: also concern about discoverer-td-identify since all discoverer should implement this
… seems that hitachi is failing on this one
… ktoumura is that fixable?

McCool: also first three features at risk are implemented by 0 implementations
… this is odd since these describe how to talk to a directory

McCool: these features do not entail automatic following of links, is more providing means to follow things to users

McCool: it seems nobody has implemented thing link

McCool: personally, if this feature is dropped it is fine, but it may bring problems if we do
… because a bunch of stuff depends on it, so we should check if dropping this may bring derived problems
… or it would be great if we can implement these, which would also solve the problem

McCool: security bootstrapping, I'm sure people is implementing these features but it is not filled in the table we should review

McCool: who is doing coap-based service

Jan: I'm dealing also with coap

McCool: that's good but we still need another one

Cristiano: I think also node-wot implements it

McCool: similar issues occur with different representations
… another odd assertion is that a successful response should return the correct content-type header
… another assertion with 0 implementations is the CoRE

Jan: we implement it

McCool: ok, we can make it informative, but it we need two implementations

McCool: ok, these were the technical ones, now let's move to the security one

McCool: we will try to request that for security we will only need one implementation
… security bootstrapping using HTTP should be easy to provide

McCool: we are out of time, but I do not feel confortable with these results
… we should update and improve the results during this one week updating the manual results
… but if we have all these at risk we will need to delay the PR

<McCool_> https://github.com/mmccool/wot-discovery/blob/oct-2022-testfest-updates/testing/report.html

Kaz: is that draft report different from the Statically version on the agenda?

<McCool_> https://w3c.github.io/wot-discovery/testing/report.html

McCool: yes, and need to update the Statically version link as well

[adjourned]