Telecon 04.04.2014
Participants
- Darko Anicic
- Avi Bernstein
- Danh Le Phuoc
- Roland Stuehmer
- Claudio Di Ciccio
- Alasdair Gray
- Jean-Paul Calbimonte
- Oscar Corcho
- Daniele Dell'Aglio
- Axel Polleres
Minutes
- JP: Continuing discussion on http://www.w3.org/community/rsp/wiki/RSP_Query_Semantics. Now looking at Roland's proposal
- JP: discussing the model, roland's proposal
- Danh: deal with streams without timestamp
- Roland: an option is to materialize the system tiemstamp
- Alasdair Gray: Distinction between Roland's and the previous suggestion is that the timestamps are embedded within the graph
- Alasdair Gray: Question is what to do about graphs without timestamps
- Roland: events should be immutable
- Alasdair Gray: Graphs should be immutable, but the processing timestamp would be something that is needed to be added later on
- from Alasdair Gray: (In this discussion, a timestamp can either be a single timepoint or an interval)
- Avi: The question is really of compactness
- Avi: If you put the intervall in the graph then you may need a graph for each event
- Alasdair Gray: Avi: whenever a timestamp is added to an event, it needs to be clearly captured in the RDF that it is an annotation on the original event
- Avi: AGree
- jp: Discussion: immutability of events, enrichment of incoming events
- Alasdair Gray: I'm happy with the capturing of the timestamps as annotations
- jp: Action: Roland to write the options mentioned and make them circulate
- Daniele: annotation of original events->links to provenance
- Daniele: what happens with events that you don't know when they end?
- Roland: in our system, don't allow events with only start time
- Roland: possible use case for this in a distributed system
- Roland: long running event pattern
- jp: Discussion: events and facts: distinction
- Axel Polleres: hmmm, very few "facts" are entirely static, or now?
- Axel Polleres: Where do you draw the line?
- Avi: Look at the publication above
- Avi: < s, p, o> [start_time, NOW() ]
- Alasdair Gray: Something like that, essentially we are allow events that do not have an end time stated
- Alasdair Gray: with the assumption that it will be added eventually
- jp: I'm fine with that possibility
- Avi: Yes!
- Daniele (PoliMI): +1
- Avi: Where is the agenda?
- Alasdair Gray: The model must be defined independently of a serialisation, i.e. it should work for all RDF serialisation. We can use a serialiation for our discussion, but the model must not depend upon it
- Daniele (PoliMI): yes, i totally agree
- jp: i agreee
- Daniele (PoliMI): my doubt is if after the model we should also talk about serialisations
- Daniele (PoliMI): or if it is out of scope
- jp: I think RDFy approaches are an advantage
- Alasdair Gray: Agenda
- Alasdair Gray: - The plan is to continue with the model discussion (hopefully we can come up with an agreement)
http://www.w3.org/community/rsp/wiki/RSP_Query_Semantics, - Short term planning: at this stage it might be good to evaluate where we are now. - Possibility of F2F at ESWC
- JeanPaul: Now talking about the possibility of starting the writing up of a report (so that we can clean up the existing bits and pieces)
- Oscar Corcho: +1
- Alasdair Gray: +1
- Danh: +1
- Oscar Corcho: I like the idea. Specially because of the cleaning up so as to avoid problems like today's email discussion
- Oscar Corcho: Avi: Proposes to agree on the document outline at ESWC
- Roland: not there, but likes the idea of the f2f meeting
- jp: +1 for doc outline
- Darko: not there either. An alternative for this could be a workshop on Web of Things (when?)
- Alasdair Gray: When and where is the Siemens event?
- Oscar Corcho: June 25th-26th
- Oscar Corcho: First f2f: working on the doc outline at ESWC
- Oscar Corcho: Second f2f: people can continue, if there, at Siemen's event
- Oscar Corcho: Who is at ESWC? Jean Paul, Oscar, Avi, Daniele, (anybody else?)
- Oscar Corcho: \methinks That sounds like a plan
- Alasdair Gray: I'm happy for those who can make it to ESWC to start drafting things
- Oscar Corcho: SIEMENS event --> W3C event on Web of Things organised by Siemens
- Oscar Corcho: ACTION (Alasdair): create a google doc or similar for everybody to help
- Oscar Corcho: Sorry, Alasdair, I assigned it to you ;-)
- Oscar Corcho: AGENDA ITEM: Next calls?
- Oscar Corcho: In 2 weeks, retake discussion on model semantics
- Oscar Corcho: assuming that there will be more invormation in the wiki
- Oscar Corcho: +1
- Darko Anicic (Siemens): +1
- Oscar Corcho: There is agreement to work on that
- Oscar Corcho: AGENDA ITEM: AOB?
- Oscar Corcho: Easter break in DERI
- Oscar Corcho: better to postpone one week
- Alasdair Gray: Shared document can be found at https://docs.google.com/document/d/1bXPD5OTebOR7SckgIyoFFB_kLAerfvI1pBgot0izO3U/edit?usp=sharing
- Oscar Corcho: next call 25th
- Oscar Corcho: next call April 25th
- Darko Anicic (Siemens): could not access the doc link
- Oscar Corcho: next calls: April 25th, May 9th