Closed dr-shorthair closed 5 months ago
Hmm. Looks like all the issues in the old repository tagged ssn
were cleaned out.
And I have not been able to track down when the Cardinalities were actually changed in the old repo.
I think I found the commit that changed the cardinalities: https://github.com/w3c/sdw/commit/882541f2b008d2935f6fb3e4becaa20901cd1e14#diff-bddd5642ff86775ffb333c45b693b0b76f4d23fbe90e3b20f4082c5dda6a8134
16th April 2017, hidden in a large set of changes to the RDF files during a big refactor by Maxime.
Looks like I was responsible for multiple hasResults: https://lists.w3.org/Archives/Public/public-sdw-wg/2017Apr/0177.html I regret that.
Can't find a record of where we agreed that actsOnProperty ≥ 1
+1 for setting cardinality of actsOnProperty
to EXACTLY 1
so it's consistent with the Observation model.
Yes, I also noticed the cardinality of MIN 1
on hasResult
(I was expecting EXACTLY 1
) and was surprised about that too... Do you recall what was the use case for having multiple results?
Another +1 for changing it (back) to EXACTLY 1
.
Does anyone know of common ontologies that would help users with multiplexing results?
I don't, but collections of actuations may contribute to the answer
In 4c36cf4 I've
PR #165
And included in superseding PR #178
Closed by #178
On Actuations, the cardinality of
actsOnProperty
andhasResult
is ≥ 1. Compare with Observations where the cardinality ofobservedProperty
andhasResult
is = 1.When we were originally formulating the Observation model 20 years ago we considered it an important principle that each Observation concerned one Property of one FoI, and generated one Result. If an application concerned complex values for any of these, then they should be encapsulated on the application side. In the general standard we did not want to get involved in implementation details, like how to map each element of a complex result to elements of a complex Observable Property, etc. That is up to the applications to manage. In the general standard we were primarily interested in standardizing the relationships and the overall shape.
However, in the 2017 edition, somehow (I clearly wasn't paying attention) we broke that pattern on Actuation. @alexrobin pointed out that this is mentioned in the changelog - item 10 in https://www.w3.org/TR/vocab-ssn/#changes-since-3rd-public-working-draft-https-www-w3-org-tr-2017-wd-vocab-ssn-20170504 I'll go look back in the GitHub issue tracker to see if there is a specific record of the discussion. I'm seriously wondering if this might be something we can unpick now.
Also see #112