Meeting minutes
Agenda
Kaz: as suggested during the Scripting, Security and Discovery calls, would suggest we as the TD-TF should quickly skim the minutes and summarize important points from TPAC meetings and generate an action item for the TD-TF and the whole WG./
Ege: ok
… (updates the agenda)
… would add slides as well
minutes
<kaz> Sep-6
Ege: it is annoying to review minutes without screen sharing
… it might be easier to have them in other formats
Kaz: we can have several options but the easiest one is to review the minutes before the call and mention the errors during the calls
Ege: right but for typos it is hard
Kaz: if we can do a task in just two minutes we can do synchronously
Ege: any problems with the minutes?
… I wasn't in the call
… ok minutes approved
<kaz> (all the speakers names have been spelled out)
next meetings
Ege: I don't see any problems with the TD call
… do you see any possible cancellations in the short future?
… ok
… no cancelations
TPAC
Ege: basically in TPAC we had the TD section on Thursday
… Sebastian presented an overview
… then I talked about the registry approach
<kaz> WoT WG/IG Day 1
<kaz> WoT WG/IG Day 2
<kaz> JSON-LD, WoT, RCH joint meeting
<kaz> APA WG
<kaz> Web Agent CG
Ege: we will do a refactoring where the Binding will be explained in the TD document
… then Sebastian introduced the different big items to deal with in the next charter
<kaz> TPAC agenda
Ege: basically a quick intro
… then I presented the binding registry
… I presented the Registry mechanism from the W3C process doc
… gave some examples
… some use the word registry but they are not really following the exact W3C Process be careful
… I talk with the co-editors of the webcodes
… good feedbacks
… but we should be careful about how the table will evolve over time
… in particular how to regulate contributions to the table
… the registry is a new concept within the W3C
… we can define minimum berries for contributing the the registry
… other than that we talk about the users and contributors
… kaz also provided other examples
… DID is one (but it is not an official registry)
Kaz: note that the w3c process provide a procedure for registry track. Any group could use the procedure. However, w3c do not provide anything on how to manage the process
… I am not sure which can be considered "true" registries because each group can interpret the process according to their view
… we can look into the best practices
Ege: I agree we have to look at our requirements and use cases
Kaz: we should mention this complication in our note
Ege: yes
Kaz: note meaning new markdown documents to be generated as McCool mentioned during the main call.
… as mentioned during TPAC, there are several WGs within W3C as well, and they also should be referred to.
Ege: Michael suggested to use a append-only approach for the registry
… we should see how it will work
… and also binding proposal should be bundled with validation
… any comments missing?
Ege: do we have an action item?
Cristiano: it could be good to have a document where we describe the policies for accepting new bindings
Ege: where do we start?
Cristiano: should we work on wot repository?
Ege: yeah but we need to take care
… of other reports
Koster: there will be a policy element
… but there are also related reports that we need to tackle. Mostly intermediate documents.
Ege: ok for using wot repository
… it can be used by other task forces
Cristiano: ok for the directory
Kaz: we as group we need to clarify the policy
… regarding the append-only approach there is some possibility of somebody requesting removing dengerous contributions (from the legal point of view)
… we need to think about it
Ege: I understand
… in any case we should add some workarounds to invalidate something
Kaz: that's why I asking to review other registries
Cristiano: +1
Koster: the whole policy should guartee a level of append-only but we should be flexible
Kaz: we should be fair to everyone too.
Luca: keep in mind that the registry contains just a description + link
Ege: the problem is the content of the linked document
Luca: is it something that you can control?
… if it is not permanent we should reject
… we don't have many ways to make something that is not legal
… it is not a big concern
… the good point to have append-only is to not breaking backward compatibility
… we are not going to host any content
Ege: the problem is when for example somebody says falsities
… we should invalidate it
Cristiano: we might need to host documents and poeple might ask to remove dangerous content
Luca: that's why we should never host those documents ourselve
… the registry is just about there is something that is accepted by enough people
Ege: ok but the point is for example matter what happens if they want to remove it
Luca: it's their fault
Ege: ok but we need to explain it
Luca: if we want to operate such a registry we should reject stuff that does not abide to such rules
Luca: if we store all the information
… of the protocol. We are open to get any kind of legal request
… if we cache the matter protocol for example and however published it was a mistake then it is a problem
… we should store the smallest unit of information
Kaz: we should discuss what we mean and what we want, that's ok (I suggested to look at other registries). For today, we should summarize the discussion for TPAC
… we can work on the detail later, as whole group.
… would like to suggest we record what we've been discussing so far as a markdown quickly.
Ege: I'll create the markdown
Kaz: Note that it's OK as an initial input for further discussion to recored what we've been discussing as a markdown. However, we should not dive into the detail on possible registry policies today during this call because we were simply summarizing the TPAC discussion.
Ege: we should followup for the action items
… maybe split the work
… we talk about this later on
Thing Description PRs
1858
<kaz> PR 1858 Remove mention of atrisk section
Ege: got the verbal approval from the director
… is it ok to merge it now?
Kaz: I had to comment on the PR before the call
<kaz> (added a comment on the approval)
Ege: ok merged
1843
<Ege> PR 1843 - Fix well known operation types only
Ege: ok it would be nice if you can comment kaz
<kaz> (gave a comment on the approval)
Ege: merged
1844
PR 1844 - Update schema to not allow empty op array
Kaz: so this is basically bug fix on JSON schema. Right?
Ege: yes
… it's about whether it's allowed to use an empty "op" array or not
Cristiano: not sure if it's correctly recorded
… but it's what we should do in the future
… you can't have an empty array
… need to address for 2.0
Luca: fine to merge it if it's consistent with the future version
Ege: (shows 5.3.1.1 Thing)
WoT Thing Description - 5.3.1.1 Thing
Kaz: in that case, do we need to merge this PR 1844 now for 1.1 version?
… or should wait until 2.0 version?
… it depends on the current status of the existing implementations
Ege: yes
Cristiano: maybe should capture the issue here
Ege: (adds two comments on the issues)
@egekorkan will check the other places that have an array value, whether they allow 0 items in the array or not.
1866
PR 1866 - TD html ontology fix|
Ege: a PR from Sebastian
… but seems he has not changed everything
Cristiano: Mahda is an expert on this
… nobody has fixed the issue yet
… we can merge this PR itself, and then can work on further fix
Ege: if we merge this, it would impact the final index.html as well...
Kaz: agree
… should be careful to fix this PR
Ege: right
Ege: (adds comments)
@relu91 will do the correct solution by changing this PR but make sure that we do not change the index.html. He will also change it for the other missing keywords as shown in TD html ontology fix #1866 (comment)
Resources directory and Namespaces
Namespace proposals
Ege: all the WoT developers will refer to this information
… so we should make sure all the resources for which version to be linked with which resource
… e.g., v1 for td-context-1.1.jsonld
… should talk about this during the main call as well
… McCool mentioned an idea on each TF's handling each resource on the TF's repository
… maybe not a bad idea
… there is a PR about this
Kaz: I still think having a consolidated area for WoT resources would be better
… even for a tentative purposes, reusing the "publishing" area under each repository would be a big confusing and embarrassing
… because that area has been used for publication preparation and is kind of messy
… so would suggest we have some more discussion about this topic, e.g., including Farshid too
Call of 20.09: The WG does not have a consensus on where to put these resources. Keeping it open until then.
Ege: would like to wrap up this topic quickly
… e.g., right after the main call
Kaz: so I suggested we have a 2-hour main call to discuss policy topics including this
Binding Templates
Note publication
Kaz: very sorry for the big delay due to TPAC and Charter procedure, etc.
… working on the document check and would like to publish it this week
Ege: ok
BACnet Binding
<Ege> wot-binding-templates PR 209 - BACnet Binding
Koster: want to check JSON-LD context file and mapping
… actual protocol verbs
… expect a context file and a mapping file. right?
Ege: outside implementers would like to have resources
… Jan has been working on CoAP
… maybe we can refer to the resources there as an example
Jan: main HTML and template for SPARQL file
… ontology documented logically
Jan: would make sense to have a guideline document
wot-binding-templates PR 282 - Initial work for vocabulary creation guide
Ege: there is a PR for a possible guideline
… but it's almost empty
Koster: need ontology.html
… is that generated automatically?
Ege: yeah
Cristiano: should have a template folder
Ege: there is an index.template.html under bindings
Cristiano: can't expect external contributors really follow the guideline
Koster: if all those files are needed, should inform what the current work flow is like
… do we require all the flow on the resources?
… what do we consider what would be the "complete" binding?
Cristiano: the idea is in the mapping, having the SHACL would be helpful
Koster: was kind of too tired to follow that...
Ege: Jan, can I ask you to add content to the proposed guideline?
… not update the PR 282 itself
<JKRhb> Sure, I will try to come up with a proposal :)
Kaz: agree we should clarify what we want
… and updating this guideline would be helpful for that purpose
… so asking Jan for help is reasonable :)
Ege: (shows the preview)
Preview of PR209 - 5. BACnet Vocabulary
Cristiano: Koster, if you have any questions, please let me know
Koster: tx!
AOB?
Ege: any other business for today?
(none)
[adjourned]