See also: IRC log
<KJanowic> so 8:00 UTC after all
<ahaller2> scribe: mlefranc
<ahaller2> scribenick: mlefranc
<ahaller2> +1
+1
<SimonCox> +1
<RaulGarciaCastro> +1
<KJanowic> +1
<DanhLePhuoc> +1
<ahaller2> https://www.w3.org/2015/spatial/track/actions/334
<KJanowic> https://www.w3.org/2015/spatial/wiki/Events_and_Situations#Events.2C_Actions.2C_and_Situations
KJanowic: oldssn:Observation
subclassof dul:Situation, now ssn:Observation is aligned with
dul:event
... dul:situation subclassof event, which is disjoint from
event
... meaning oldssn:Observation equivalent with ssn:Observation
cannot hold
... so where do we break ?
... we decided that oldssn would be where we break,
... because ssn is going to be the standard, and withstand
longer
... there are now more little subtelties
... for example platform
... about the predicates: sixth bullet point, ...
... alignement of properties with associatedWith don't bring
anything, except everything becomes instance of
dul:Entity
... solution proposed: we can used general concepts and
intuition axioms: we avoid unguarded domain and range
restrictions
... so although we can't say oldssn:Observation equivalent with
ssn:Observation,
... we could say: for all those things that have some
sosa:hasResult property and is not a oldssn:Observation, then
it is a sosa:Observation
... so we create local alignments
<ahaller2> +1 for KJanowic axioms!
<KJanowic> and Result and Deployment (which are not covered or broken, respectively)
ahaller2: in favour to add local alignments
<SimonCox> Is it possible that the difficulty with the DUL alignment is (a) a misalignment in the original SSN-DUL design, or (b) an underlying difficulty in DUL?
<KJanowic> This is the axiom: \exists old-ssn: observationResult.T \sqcap \not old-ssn:Observation \sqsubseteq SOSA:Observation
mlefranc: would like to see implementation
<SimonCox> If either of these is true, then I am not so interested in DUL 'getting in the way' of SSN/SOSA moving forward ...
KJanowic: because we don't have
domain and range axioms for properties, we can say: if someone
uses a property that links x to y
... and x is not asserted to be a oldssn: something,
... then we could align x and y to the proper dul class
* please check my minutes
KJanowic: would be a partial
alignment
... users of oldssn would need to remove part of their
triples
<KJanowic> +1 to that!!!
ahaller2: we can have a note in the old observation class that says: instances of oldssn:Observation should migrate to ssn:Observation
mlefranc: there may be side
effects, we will have to wait for the wide review.
... don't really know actual oldssn implementations that
heavily rely on the alignment to dul
DanhLePhuoc: I don't know if we
addressed all the communities that have used ssn so far
... we need to wait for people to speak up
<DanhLePhuoc> +q
KJanowic: I think it's better to
help people to migrate
... three options for alignment: 1 no alignemnt 2 partial
alignment, 3 migration
<SimonCox> so KJanowic suggest that it is OK to say 'these parts can be aligned, but those can't'
ahaller2: if we don't align to dul anymore,
<KJanowic> I agree!
<KJanowic> Yes, this is why I favor breaking the old not the new
ahaller2: then for people that use the old ssn - dul alignment, it may happen that instance of new ssn:Observation becomes instance of dul:Situation
<KJanowic> +1 to Danh
<ahaller2> +1 to some note to tell you how to migrate
DanhLePhuoc: let's do the
alignment that is safe, and write a note about the gray
area
... current alignment covers 70% of terms,
<KJanowic> This is exactly the partial alignment that the axiom does
DanhLePhuoc: for the rest, people can use rules, or do something else
+1
<KJanowic> Okay
<ahaller2> ACTION: Implement the proposed axiom to partially align observationResults of oldSSN to new SSN in the ontology files [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action01]
<trackbot> Error finding 'Implement'. You can review and register nicknames at <http://www.w3.org/2015/spatial/track/users>.
<ahaller2> https://www.w3.org/2015/spatial/track/actions/335
ahaller2: removed that from the
document, wanted to discuss that here
... few mistakes spotted after the document was published in
the TR space
<ahaller2> close ACTION-335
<trackbot> Closed ACTION-335.
ahaller2: about figures for instance
KJanowic: aren't all alignemnt non-normative ?
ahaller2: yes, but the one to
o&m
... o&m implementation may become ssn implementation
evidence
... the graphs must represent that
KJanowic: but Phil told us that may not help to convince the director
ahaller2: that's a gray area
SimonCox: I can't advise on that
ahaller2: we need Phil and Francois to help here
SimonCox: at least things that are marked at risk can be dropped so that we can drop them later without going back to candidaterec track
mlefranc: is it a good idea to have no examples ? recent discussions show that examples are crucial
ahaller2: examples may be incorporated lated, the goal here is to freeze the change list
KJanowic: asking SimonCox about previous features at risk
<ahaller2> close Action 334
<ahaller2> close action-334
<trackbot> Closed action-334.
close ACTION-334
<trackbot> Closed ACTION-334.
SimonCox: new classes and
properties in new OWL time ontology have no implementation
yet,
... so we needed to mark those that have been discussed and
introduced the latest: "at risk"
... to avoid that the whole documents fails passing the
implementation evidence test
<KJanowic> thanks a lot!
ahaller2: working draft will be early May, so we can start identifying classes and properties that are at risk
RaulGarciaCastro: if ssn:qualityOfObservation is marked at risk and has no implementation evidence, will it be removed, or become non-normative ?
SimonCox: bits of the ontology may be non-normative
ahaller2: if it's confusing in the working draft, why would you include them in the ontology ?
SimonCox: or provide a second ontology that includes non-normative aixoms ?
ahaller2: yes
mlefranc: let's be optimistic here, if some term that is marked at risk has no implementation evidence, we might be able to move them in a "non-normative" section
<DanhLePhuoc> +q
ahaller2: we may mark some term
at risk because we did not have enough time to discuss
them
... we will have to ask Phil and Francois about if we need to
remove terms that are dropped from the ontology
KJanowic: showing that we take
wide review comments into account may be actually more
important that we supposed yet
... we need to have more reviews, and show the Director that we
took them into account
<KJanowic> sorry for jumping ahead
ahaller2: maybe send emails to
all ssn users individually
... for example recent email triggered issue https://www.w3.org/2015/spatial/track/issues/187
... it's very important that we address these issues as fast as
possible
DanhLePhuoc: I sent review requests to iot.schema.org,
<DanhLePhuoc> https://groups.google.com/forum/#!topic/sdo-iot-sync/132-dZf0kTU
DanhLePhuoc: and WoT-LD taskforce. Which is very crowded, I hope we will receive reviews soon
ahaller2: of course reviewing takes some time
DanhLePhuoc: in terms of
implementation, if a new concept has a
equivalentClass/equivalentProperty to the old one,
... then old ssn data that used that concept will become
implementation evidence for the new term
ahaller2: we will just mark classes and properties at risk if they had no impelmentation evidence
<ahaller2> ACTION: ahaller2 to mark classes/properties as at risk in the ED [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action02]
<trackbot> Created ACTION-346 - Mark classes/properties as at risk in the ed [on Armin Haller - due 2017-05-09].
ahaller2: I will send email to original users of SSN
<ahaller2> ACTION: ahaller2 to invite original users of SSN via personal email to review the document [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action03]
<trackbot> Created ACTION-347 - Invite original users of ssn via personal email to review the document [on Armin Haller - due 2017-05-09].
ahaller2: evidence that we covered public review comments is important
<ahaller2> https://hal.archives-ouvertes.fr/hal-01467853
ahaller2: the SAN ontology is
parallel to core SSO in SSN, but for actuators
... we're ok to introduce effect class ?
<KJanowic> btw,weused san as input for out actuator work in SOSA
KJanowic: we used that actuator
pattern, we have proof for that in the wiki
... do we really want to introduce Effect now, we know that
Stimulus can't have instances
... every term needs implementation evidence or discussion to
explain why they can't have
<KJanowic> okay
KJanowic: if we don't need to have implementation evidence, then ok to introduce them. Else, no
<KJanowic> Can you asign an action to me? also for the dul axiomatization?
<ahaller2> ACTION: KJanowic to send email to Phil and Francois to clarify the need for implementation evidence for superclasses and documentation classes [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action04]
<trackbot> Created ACTION-348 - Send email to phil and francois to clarify the need for implementation evidence for superclasses and documentation classes [on Krzysztof Janowicz - due 2017-05-09].
Fig 2. in https://hal.archives-ouvertes.fr/hal-01467853/document ---> Sensor produces Result vs Actuator consumes Input
<KJanowic> IMHO, it is too late to introduce major changes. we need to move on
<KJanowic> [I have to leave now]
<ahaller2> scribe: roba
<ahaller2> scribenick: roba
<mlefranc> https://www.w3.org/2015/spatial/wiki/Linking_Systems_and_Features_of_Interest
RaulGarcioCastro: foi is related
to system only via Observtion class.
... proposal is to move property to SOSA to allow sensors to be
related to features without Observation
mlefranc: question is whether
observableProperty is a "general" property such as speed, or a
feature property :"speed of this car"
... withous hasProperty SOSA cannot support this,
armin: agree - users of SOSA must use the spedific property as documented. This doesnt solve the competency question
<SimonCox> http://registry.it.csiro.au/def/environment/_property
armin: naming of properties is "pretty bad"
SimonCox: when looking at sets of properties - are they individuals or properties of classes? QUDT has both.. bigger debate than here
<RaulGarciaCastro> It is not about moving Property
<RaulGarciaCastro> Property can stay at SSN
SimonCox: re move Property to
SOSA - judgement re weight of SOSA
... too much of a change at this point
... type of property without knowing the feature of interest -
but can also use features such as "the World" - properties
cannot exist without Features in the OGC reference model.
mlefranc: O&M and SSN were well aligned - the only question is whether hasProperty and isProperty to SOSA would enable SOSA to cover some important cases
ahaller2: discussion was these names were confusing - would prefer new name
mlefranc: ok to rename, "quality" would be perceived as disjoint from "quantity" - first step is to move, second step is to rename
<SimonCox> BTW - we have some implementation evidence on SOSA samples, sampling, sampler (and sample-sample relationships!) emerging by our friends at Geoscience Australia
ahaller2: action item to create a PR - to Raul. - for discussion
<ahaller2> ACTION: RaulGarciaCastro to issues pull request to move isPropertyOf hasProperty to SOSA, and propose different names/labels [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action05]
<trackbot> Created ACTION-349 - Issues pull request to move ispropertyof hasproperty to sosa, and propose different names/labels [on Raúl García Castro - due 2017-05-09].
discussion of examples - updating PR against feedback (please paste in link to the feddback being discussed)
<ahaller2> ACTION: mlefranc to aggregate his examples from an earlier pull request and introduce them in the beginning of each section with the Observation/Actuation/Sampling figures [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action06]
<trackbot> Created ACTION-350 - Aggregate his examples from an earlier pull request and introduce them in the beginning of each section with the observation/actuation/sampling figures [on Maxime Lefrançois - due 2017-05-09].
ahaller2: who do we know is doing an implementation?
SimonCox: will help GA by review. After Review will be update of an operational system.
<ahaller2> ACTION: Simon to collaborate with Nicholas Car from Geoscience Australia on the SOSA implementation in their operational system [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action07]
<trackbot> Created ACTION-351 - Collaborate with nicholas car from geoscience australia on the sosa implementation in their operational system [on Simon Cox - due 2017-05-09].
DanhLePhuoc: will implement one NOAA dataset
<DanhLePhuoc> NOAA
<ahaller2> ACTION: DanhLePhuoc to coordinate one implementation evidence around SOSA/SSN in the NOAA dataset [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action08]
<trackbot> Error finding 'DanhLePhuoc'. You can review and register nicknames at <http://www.w3.org/2015/spatial/track/users>.
DanhLePhuoc: 20K+ weather observations
ahaller2: will update historical
climate dataset - small subset of terms
... timeline is about 6 weeks
<ahaller2> ACTION: ahaller2 to implement SSN new in lab.environment.data.gov.au [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action09]
<trackbot> Created ACTION-352 - Implement ssn new in lab.environment.data.gov.au [on Armin Haller - due 2017-05-09].
ClausStadler: not sure if my project will be available in time
<mlefranc> Modules to be covered once or twice: actuations / input-output / systems and subsystems / platforms and deployments / systemcapabilities, operatingranges, survivalranges
<ahaller2> ACTION: DanhLePhuoc to talk to people in Siemens to have an implementation in one of their Plugfests [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action10]
<trackbot> Error finding 'DanhLePhuoc'. You can review and register nicknames at <http://www.w3.org/2015/spatial/track/users>.
<ahaller2> ACTION: danh to talk to people in Siemens to have an implementation in one of their Plugfests [recorded in http://www.w3.org/2017/05/02-sdwssn-minutes.html#action11]
<trackbot> Created ACTION-353 - Talk to people in siemens to have an implementation in one of their plugfests [on Danh Le Phuoc - due 2017-05-09].
DanhLePhuoc: will explore possibilities in ioT plugfests
<RaulGarciaCastro> Bye!
<mlefranc> bye
This is scribe.perl Revision: 1.152 of Date: 2017/02/06 11:04:15 Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/ Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00) Succeeded: s/renames/rename/ Present: ahaller2 mlefranc KJanowic DanhLePhuoc SimonCox RaulGarciaCastro ClausStadler ScottSimmons Found Scribe: mlefranc Found ScribeNick: mlefranc Found Scribe: roba Found ScribeNick: roba Scribes: mlefranc, roba ScribeNicks: mlefranc, roba WARNING: No meeting chair found! You should specify the meeting chair like this: <dbooth> Chair: dbooth Found Date: 02 May 2017 Guessing minutes URL: http://www.w3.org/2017/05/02-sdwssn-minutes.html People with action items: ahaller2 danh danhlephuoc implement kjanowic mlefranc raulgarciacastro simon[End of scribe.perl diagnostic output]