Meeting minutes
Minutes Review
<kaz> Mar-19
Luca: any changes to the minutes?
Luca: are we ok to approve?
Luca: minutes are approved
Actions in Profile
Issue 369 - Clarifications on async actions
PR 266 - Refine sync vs. async action protocol binding - closes #259
Luca: Can you find some issues that are about this
Ege: Here are two
Luca: it is under specified in the TD as well
Kaz: the details of the actions should be specified by the TD first and profile should refer to it
Luca: TD doesn't describe anything at the moment
Kaz: We should raise issue in the TD first in that case
Luca: cannot be done in TD 1.1. Should be in TD 2.0.
… do we want to patch the profile?
Luca: TD is underspecified but an async action is already an unknown behavior in the TD consumers
Ege: so the question here is that the description of TD 1.1 is not clear enough. right?
Luca: yes, that is the question now
Ege: think the output schema for TD can be fixed
Ege: One part is that we do not have examples of TDs for different styles of actions. So we do not have an agreement on how a TD should look like for those and the discussions are not grounded
… Also the profile is not a subset of TD and is breaking implementations like node-wot but any implementation that is strictly TD based will have issues adapting to profile
Luca: if we cannot have an agreement on how actions should look like, we can just remove it but the spec becomes less useful
Ben: There needs be a WG decision on whether profiles can go beyond what is specified in TD for parts where the TD is not expressive enough
… this is the case for actions but also for SSE
Luca: then we should have a main call resolution
Ben: that is the main blocker
<Zakim> Ben, you wanted to react to Ege
Kaz: In short I agree with Ben
Kaz: If TD is underspecified, it is the issue of TD and that spec should be improved
… in that case profile TF should decide what the Profile spec should contain
… also we should decide whether the current draft should become a REC or not
… I mentioned in the previous call that I think that current drat is not ready to become a REC. So we should clarify what to be described by TD 2.0 and what to be described by Profile, and those specs can become RECs for this Charter period. That updated Profile still can be a WoT Profile 1.0.
Luca: I would like to move to the next topics in the next 20 mins
Ege: I do not see implementations of the current profile in the oss projects I am involved in or within Siemens. So I don't see any issues going directly to 2.0 but name it already 1.0
Ege: regarding going beyond TD, when we do that (which we do), an implementation like node-wot has to have another codebase per profile or have a lot of if statements
Luca: in my implementation experience, profile is just a set of bindings
Luca: (Ege could not take correct minutes for the next points)
Luca: we should explain the pain points
… if we cannot solve pain points, we can move to 2.0 and make 1.0 a note only
Ben: There are complicated details to specify in a TD. Binding Templates needs more defaults as well
… before we having to do that in the profiles
<Zakim> Luca, you wanted to react to Ege
Ege: We need to decide whether we want 1.0 as REC or not. From what I see here, you are not wishing that at the moment?
Luca: I think it is fine to not have REC but we should close it asap
Kaz: the main call tomorrow is quite full
… but we can have a 5 min discussion starter or heads-up
<kaz> [adjourned]