<scribe> scribenick: kaz
McCool: goes through the agenda on the wiki
McCool: updated the draft
... pending PR 294
McCool: will merge it tomorrow after
minor corrections
... including implementations from Intel, Siemens,
Panasonic
... fair number of implementations
... 7.1 result table
... MUSTs and MAYs
... collect data on how many implementations implemented each
feature
... marking features without enough implementations
... general assertions all the vocabularies are required
... id, vocab-base, etc.
... parent assertion (general ones)
... and textual things on syntactic context
... that's the dependency
... and then
... interoperability test results
... at 7.2
... Intel's OCF metadata translator, etc. vs other
implementations
... description under /testing/inputs/interop/
subdirectory
... e.g., intel.csv
https://github.com/mmccool/wot-thing-description/tree/updated-test-results/testing/inputs/interop
McCool: for CR/PR transition, we
don't need this part
... but would be useful for implementers to check
interoperability
McCool: all the implementers need to
generate implementation reports like the above
... there is another file
... can do this table imple.csv
... each implementer to submit an HTML file
McCool: and results
https://github.com/mmccool/wot-thing-description/tree/updated-test-results/testing/inputs/results
Kaz: so 2 files?
... HTML and result.csv?
McCool: result.csv can be done during the TestFest
McCool: can work with the table
rendered by GitHub
... need description on the description about each
implementer's implementation
... as HTML
Kaz: that description will go to the pinkish area within the plan.html
McCool: right
... need to merge the PR
... tomorrow
... meanwhile please submit your implementation reports to my
repo
McCool: any questions?
dsr: wondering about adding a link to the WG page?
McCool: this is still draft
Kaz: after McCool's merging this
to the wot-thing-description repo, we can add some description
and links to the wot-thing-description repo
... all the PlugFest participants are encouraged to add their
information
... to your repo today
... or to wot-thing-description repo after tomorrow
McCool: will merge the PR within 12
hours
... or you can send your report to me by email
... or via PR as usual
<McCool> https://github.com/w3c/wot/tree/master/plugfest/2018-lyon
<kaz> Fujitsu's PlugFest report summarizing the results
McCool: (refering to Fujitsu's PlugFest report)
... not sure who was connected to Intel
... we have these checking points
... validation, for example
... ideally we can talk about validation suite
... what was validated
Ege: errors? or just what was validated?
McCool: we can generate result CSV files as a side effect of validation
Ege: some sort of JSON
schema?
... to identify which assertion
McCool: generate CSV files
... we can generate this as a side effect of playground
Ege: ok
McCool: another discussion here
... 4 statuses: pass, not-imple, failed, not mention
... I count "not-imple"
... means TD has no actions, possibly
Ege: could mean if you can't have
2 things at once
... either property or action
McCool: the way I'm counting the
implementations
... chunks of codes
... if someone has more than one implementation
... need some kind of logical merge
Ege: sometimes not all the TDs can't be supported at once
McCool: let me try some result merger
Ege: will generate CSV per TD
McCool: ok
... go back to Fujitsu's plugfest report
... 3.1.3 Connect with Remote/Local Proxy
... Thing Directory consumes the TD and reproduces it
... manipulate the TD
... still counts as an implementation experience
... here Term definition within criteria.md
<kaz> https://github.com/w3c/wot/blob/master/testing/criteria.md
McCool: section on what to be
included
... and what not to be included
Ege: what test bench?
McCool: node-wot could be a component
<kaz> Fujitsu's PlugFest report summarizing the results
McCool: go back to the plugfest
report again
... 3.1.2 register with Thing Directory
... need to clarify the relationship
... between the plugfest report and implementation report
McCool: week of Dec 10
... 2 hours on Monday (Scripting/Security)
... 3 hours on Wednesday (Editor/Main/Testing)
toru: what time?
McCool: 9pm on Monday
... and 9pm on Wednesday
... and then 4pm on Friday
... all JST
... create a wiki page
<scribe> ACTION: kaz to create a wiki page for the TestFest
McCool: 2 weeks from now
... feasible?
... in the worst case, we can collect TDs
... and fill out the result files
... validator to create the result files?
Ege: can work on it
McCool: we can collect TDs
... generate subset of assertion result CSV files
... if we could have an automated tool (based on the
playground) that would be great
... if not, we can do that manually
... also we can collect interop.csv as well
... and we can update the draft implementation report
... note that we still need implementation.html description
from implementers
... also still useful to have human-readable
preparation.md/results.md
... btw
... still working on VPN server setting
... based on an OpenVPN server
... anything else?
... wondering about how to allocate a 3-hour call
Kaz: we can simply extend the first hour, e.g., the Editor's slot
McCool: would like kaz to include a template including the links for the calls
McCool: within the plan.html
... test spec setting in appendix
... smaller team to work on that?
Kaz: if people are interested they can join. right?
McCool: me, Ege and Dave are the initial candidates
taki: interested
... if additional assertions to be added
Kaz: as well :)
<McCool> https://www.w3.org/WoT/IG/wiki/WG_Binding_Templates_WebConf
McCool: binding template call on Dec. 4?
Kaz: have a dedicated discussion
with Ralph on testing requirements at that time
... so how about having the discussion tomorrow on Nov 29 at 11pm JST (6am US Pacific, 9am US Eastern, 2pm UK, 3pm EU)?
all: sounds good
Toru: when to fix the TD spec for the TestFest?
taki: i18n aspect, etc.
... a couple of items to be done
... all properties at once, etc.
... to be discussed on 30th
McCool: backward compatibility with
Lyon version?
... thing template, etc., are optional
... don't remember anything breaks backward compatibility
... would focus on Lyon TDs
... but also would generate assertions based on the latest
draft
Kaz: ok with that direction
... the results from the TestFest don't have to be
perfect
... Taki, please note that we'll have a W3C Japanese member
meeting on Nov. 30 and many of the JP participants including
myself can't make the TD call on Nov. 30
[adjourned]