RE: [rsp] work in progress

Thanks Alessandra, Roland

@Roland: true, what you mention goes beyond what has been discussed in serialisation, and it's very important. One thing I noticed is that when we discuss about defining 'Events', there is a bit of confusion. Sometimes we think of them as high level domain events (i.e. a possible use case of RSP), but you refer to more fundamental questions, and refering to events as the core abstraction of complex event processing.

It seems to me that some of these important issues fit well with the general RSP model discussion (e.g. punctuation, model events in graphs, etc). Of course all of this will have an impact in the serialisation discussion, that's why I linked the two.
best,
Jean-Paul 


From: alessandra.mileo@insight-centre.org
Date: Wed, 5 Nov 2014 10:26:40 +0000
CC: public-rsp@w3.org
To: jpcalbimonte@delicias.dia.fi.upm.es
Subject: Re: [rsp] work in progress

Dear Jean Paul, all,I regret I didn't make it to ISWC, I just came back from leave 10 days ago, still catching up :)
A note regarding proposal 2.I remember we had a discussion around the issue during our meeting in Berlin.There was a clear interest on my side and Adrian's to understand the scope of more complex logic within RSP processing, but we agreed that with this state of things it would be probably more in scope if we create a different task force (driven by Adrian/Kia, myself and Darko), to discuss how more complex logic (and related operators) can fit with the RSP semantics.Since such agreed semantics will be main focus for next year, I believe it would be good if this task force I am proposing goes along but without interfering, at least not at this stage, with objectives 1,3 and 4.Maybe whoever added 2 can comment further?
Best,Alessandra

-----------------------------------------------------------
Dr. Alessandra Mileo
Research Fellow & Project Leader
INSIGHT Research Center
NUI Galway, Ireland
Email: alessandra.mileo@insight-centre.org
-----------------------------------------------------------




On 05/nov/2014, at 09:34, Jean Paul Calbimonte <jpcalbimonte@delicias.dia.fi.upm.es> wrote:Dear all,

As said in last friday's call, we are still receiving comments for editing the Group Note:
https://docs.google.com/document/d/1bXPD5OTebOR7SckgIyoFFB_kLAerfvI1pBgot0izO3U

We have made a large number of changes, including general editing and integration of the query features wiki page. This last part needs a bit more consolidation. Please feel free to edit and/or comment on the doc or if you want to share something in this list, do it as well.

The other main point was the selection of tasks for the 2nd year. According to Avi's braninstorming tool, we have the top 5 following things:
Agree on query evaluation semantics71Structuring complex logic that a RSP engine could process (design patterns, tools, ...)69Syntax for RSP query language67Semantics of RSP language66Propose reference event model (well-aligned with existing time ontologies, supporting different time models, supporting serialisation, )

There is clearly a large interest in agreeing on an RSP language (1,3,4). So if you all agree this would be our main task. As discussed last friday we'd like to do it iteratively: we can agree on a basic query language and then preogress adding more operators/features.

As for the other suggestions (2 & 5). I do not fully understand proposal 2. Can you (those who proposed it) elaborate a bit more? 
About proposal 5 it is linked to serialisation and usage of well defined vocabularies (e.g. owl tie, prov) for representing stream items, if I understand right. Then this can be undertaken in depth by the serialisation taskforce that has already started within the RSP group.

Cheers from rainy Lausanne,

Jean-Paul
 		 	   		  

Received on Wednesday, 5 November 2014 15:06:26 UTC