RIF] Draft of minutes from June 27th telecon

Dear RIFWG,

Here's a draft of the minutes from yesterday's telcon, for your review. 
Thank you to all who helped fill in the scribing gaps in the IRC.

Stella





- DRAFT -
RIF WG weekly telecon minutes
27 Jun 2006
See also: IRC log
Attendees
Present
Allen Ginsberg, Christopher Welty, Christian de Sainte Marie (csma), Dave 
Reynolds, David Hirtle, François Bry, Harold Boley, Jeff Pan, Jos de 
Bruijn, Leora Morgenstern, Markus Krötzsch, Mike Dean, Paula-Lavinia 
Patranjan, Sandro Hawke, Stella Mitchell 
Regrets
Evan Wallace(OMG meeting), Frank McCabe (Also at OMG), Igor Mozetic, 
Michael Kifer, Michael Sintek (project meeting), Peter F. Patel-Schneider
Chair
Chris Welty
Scribe
Stella Mitchell
Contents
Topics 
1.      ADMIN
2.      F2F Meetings
3.      Liason activity
4.      Use Cases and Requirements
Summary of Action Items

 
 
ADMIN
ChrisW: The next telecon is on July 4th. We will vote on releasing the 2nd 
draft of the UCR document. ... If you have concerns with releasing it, you 
should let the group know via email by Friday, June 30th. If there are any 
concerns, the vote will be postponed. 
<ChrisW> 
http://lists.w3.org/Archives/Public/public-rif-wg/2006Jun/att-0146/20-rif-minutes.html
<ChrisW> last week's minutes
ChrisW: Propose to accept the minutes of the June 20th telecon. 
Objections? ... None. Approved. ... Any proposed ammendments to today's 
agenda? ... None.
... The minutes from F2F3 are almost complete.
Christian: I will complete the remaining minutes from F2F3 by June 29th.
<ChrisW> http://www.w3.org/2005/rules/wg/wiki/F2F3#preview
<ChrisW> http://www.w3.org/2005/rules/wg/wiki/F2F3
ChrisW: The minutes from F2F3 are linked into the agenda on the wiki page 
for the F2F. ... The action list from F2F3 is up to date.
... people should review the minutes
F2F Meetings
<ChrisW> Peter's comments on f2f4:
http://lists.w3.org/Archives/Public/public-rif-wg/2006Jun/0139.html
Christian: Peter PS is not here, but sent email with information about 
F2F4.
ChrisW: The contracts have been signed, and we hope that registration will 
be open by the end of July.
ChrisW: As discussed at F2F3 in Budva, we plan to have F2F5 in early 2007. 
Please think about hosting F2F5. Our preference is to have it in the US or 
Canada since the last two meetings were in Europe, but since there have 
been no proposals so far, anything is welcome.
Liason activity
ChrisW: We need a liason for SPARQL (W3C)
<josb> we are not in sparql
ChrisW: Something needs to be done there.
Christian: The OMG PRR group was supposed to have a F2F meeting today, but 
no one showed up except me. We submitted a revised draft to OMG and plan 
to submit the final draft by mid-December.
<sandro> ChrisW, did you want to mention the Common Logic draft, mentioned 
on the mailing list.... ? 
Christian: The OMG PRR draft will be submitted to RIF members for review, 
provided that ILOG, IBM and Fair Issac agree.
<sandro> Christian: If ILOG, FI. and IBM agree, the PRR draft can be 
circulated to RIF participants --- since the submitters & OMG agree, they 
can redistribute to whoever you want.
<scribe> ACTION: ChrisW will check that releasing PRR draft to RIF members 
is ok with IBM [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action01]
ChrisW: Christian and Chris have an outstanding action to come up with a 
process for deciding on UCR text about "cover." They are working on this. 
<ChrisW> http://common-logic.org/
ChrisW: Chris Menzel sent the final technical draft of Common Logic (ISO). 
The URL is included above. I believe that unlike many ISO standards, this 
one will be freely available.
ChrisW: ODM (OMG) is in a similar state to Common Logic, i.e. they are 
waiting for implementation experience before putting the final stamp on 
it. ...Evan W (a liason for ODM) is not on the call.
Use Cases and Requirements
<DavidHirtle> draft:
<DavidHirtle> http://www.w3.org/2005/rules/wg/ucr/draft-20060626
ChrisW: Above is the link to the 2nd draft of the UCR document. David H, 
can you talk about this draft?
DavidH: There are still some outstanding issues. But it would be good to 
get some feedback from non-editors. We haven't heard from any non-editors 
yet.
<PaulaP> +1
ChrisW: As previously decided, people have until June 30 to comment
DavidH: Peter PS made the point that there are some problems and 
inconsistencies in the 'motivates' links (from use cases to requirements) 
in the use cases section.
DaveR: Regarding the "motivates" links -- I suggest that we agree on a 
consistent set, or else remove them from this draft
Christian: We need to have a CSF that addresses alignment with relevant 
standards -- ALL the the relevant standards -- and that CSF will support 
the widescale adoption goal, and then another CSF which is 'consistency 
with key W3C specifications,' and this second CSF is the one that pertains 
to the Semantic Web.
DaveR: I don't disagree; but that is not what is within the current 
document
<DavidHirtle> it's shown in this diagram:
<DavidHirtle> 
http://www.w3.org/2005/rules/wg/wiki/UCR/Goals?action=AttachFile&do=get&target=Goals.png
DaveR: There are two alignment CSF's in the current document: 'Alignment 
with key W3C Specifications' and 'Alignment with Semantic Web'
ChrisW: What is Christian's point?
Christian: I agree that we should merge the two Alignment-related CSF's 
mentioned above, but I think we should also add an additional one which is 
'Align with all relevant standards'
DavidH: We don't have any requirements for alignment with widely deployed 
standards (in general) so why have a CSF for it?
Christian: The 'Support XML' requirement supports the 'widescale adoption' 
goal, so we need to have an XML syntax. If we move 'support XML' to 
'alignment with semantic web', then I agree
ChrisW: In summary, one proposal was to merge the current two alignment 
CSF's, and if we merge them, then Christian said we need a new CSF. 
However, DaveR said that if we add the new CSF there would be no 
requirements for it. ... I notice that the text and picture in the 
'Goals/Requirements/CSF' section don't agree.
DavidH: Frank McCabe changed the picture yesterday. The diagram is more up 
to date than the text.
DavidH: We don't need 'alignment with other specifications (in general)' 
because there are no requirements for such alignment
Christian: I think that in some circles, just saying RIF will be XML-based 
will not be adequate
DaveR: Re: XML-based - Do we have a missing requirement?
Christian: Yes, we didn't add in this requirement from the Charter
ChrisW: (question about 'support alignment with widely deployed 
standards')
... is there an objection to the way the diagram is now?
DavidH: Yes, it is inconsistent.
ChrisW: and the merged CSF would support 'W3C Consistency' goal as well
Christian: I would say it supports only 'Widescale adoption'
ChrisW: I understand Christian's point that the widescale adoption goal is 
not just for XML
Christian: Merge 'Alignment with Semantic Web' into 'Alignment with key 
W3C specifications'
... if we can only keep one, we want to keep 'key W3C specifications'
ChrisW: Any objections?
... the name of the merged alignment CSF will be 'Alignment with key W3C 
specifications'
<DavidHirtle> proposal: one CSF called "Alignment with key W3C 
specifications"
DavidH: Agreed. And later we can add another CSF if we need to
<csma> ack
<DavidHirtle> proposal: RDF, OWL and XML requirements support this CSF, 
and this CSF supports both "Widescale adoption" and "W3C Consistency"
DaveR: I am not completely comfortable with this proposal, but am not sure 
how to phrase my reservations... ...not mentioning the Semantic Web seems 
like a slight step backwards
ChrisW: If we expand the text for the CSF to mention the Semantic Web, 
would that satisfy you?
DaveR: Yes, that would be better.
Christian: Better to have it in the text because it may be confusing for 
non-Semantic-Web people if we leave it in the title
ChrisW: And also, other specs such as xquery, etc may also be key, so we 
want the phrasing of the title to indicate that we are still considering 
them.
ChrisW: Speaking of XML syntax, that's one of the items in Sandro's email 
about 'Input to UCR from Charter'. Sandro, Leora, and DavidH analyzed the 
requirements section of the UCR document to see whether anything from the 
WG charter is missing.
Leora: A number of requirements from the charter are not explicitly 
mentioned in the requirements section because it is assumed that they will 
be covered in the RIFRAF.
ChrisW: Yes, at F2F3, we discussed that we will postpone decision about 
some requirements until the RIFRAF is expressive enough to capture them.
Sandro: Someone just skimming the UCR document might miss the 'coverage' 
link at the beginning of section 5 that goes to the UCR/coverage section 
in the wiki.
<sandro> so let's add something to the introduction of the Requirements 
sections to say something like "PLEASE NOTE: Many requirements will appear 
in the _Coverage_ section, and not here"
Christian: (agrees)
<scribe> ACTION: Leora will add text to introduction - will change it in 
the wiki [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action02]
<sandro> introduction of Reqiuirements section, that is.
<DavidHirtle> (Sandro's email re: Input to UCR from Charter)
<DavidHirtle> 
http://lists.w3.org/Archives/Public/public-rif-wg/2006Jun/0127.html
ChrisW: DavidH posted the URL of Sandro's email message. Let's review the 
items listed in the email.
... First, is SPARQL a phase 1 requirement?
Sandro: Yes, I think it is.
ChrisW: Does anyone object to 'compatibility with SPARQL' being a phase 1 
requirement?
Sandro: Note that this requirement is not about the capability to make 
external calls from within a rule, rather it is about SPARQL as a language 
for query of the dataset.
Christian: Perhaps we should add a statement to the requirements section 
of the UCR document to say that all requirements from the charter apply.
DaveR: I want to return to Sandro's point about the wording of the charter 
regarding the use of SPARQL. Depending on how you interpret 'data set' I'm 
not sure it's clear that it doesn't refer to external queries
ChrisW: We need to postpone this discussion re: SPARQL and put it on a 
future agenda. ... Second item under first topic from Sandro's email: XML 
syntax. Does anyone object to this requirement? ... No objections.
<sandro> " The primary normative syntax of the language must be an XML 
syntax. "
<sandro> RESOLVED
<scribe> ACTION: David H will add the XML req to ? [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action03]
(to the UCR document) 
ChrisW: Next topic from Sandro's email: ...Regarding the 'Extensibility' 
CSF, what did we decide at F2F3?
sandro: It's a CSF and we decided not to discuss CSFs at F2F3.
Christian: There is a CSF called 'Extensibility' ...
<sandro> http://www.w3.org/2005/rules/wg/ucr/draft-20060626#Extensibility
ChrisW: The 'limited number of dialects' requirement also supports the 
'Extensibility' CSF.
sandro: No, support goes the other way
<DavidHirtle> anyone have problems with following wording?
<DavidHirtle> "RIF must have an XML syntax as its primary normative 
syntax."
<sandro> Compliance Model supports CSF:Extensibility
DaveR: I don't see the difference between 'Compliance model' and 'Default 
behavior.'
Sandro: There is a difference:
<sandro> csma: Compliance Model :: you must be able to be compliant 
without supporting everything in RIF
<sandro> csma: Default Behavior :: This supports predictibility -- it says 
what you do when you encounter something you don't support.
Sandro: Can someone clarify the text? You couldn't have default behavior 
without a compliance model
DaveR: Agreed
<sandro> Note that Compliance Model is implied by Default Behavior. 
Arguably it's such a necessary/obvious requirement it doesn't need to be 
written down.
ChrisW: Next item from the email: UCR for OWL
sandro: It seems that this is a CSF for RIF, and it is in our charter
<sandro> (agreement from DaveR)
ChrisW: Is it already covered by the 'coverage' or the newly merged 
alignment CSF?
Sandro: The alignment CSF is more abstract
Leora: This isn't a requirement per se, it didn't belong in that part of 
the document
Sandro: It is a CSF.
ChrisW: But are the SPARQL and OWL items already captured in the current 
document under the 'coverage' and 'alignment' CSF's?
Sandro: No, I think it's different. Some OWL users were upset because they 
felt that we were coming up with competing technology that would not be 
easy to use together with OWL.
ChrisW: 'Easy to use together' is difficult to measure
Sandro: This might support some of the requirements about RDF data and OWL 
data
<PaulaP> OWL data is a Phase 1 requirement and I think this is enough for 
the moment
ChrisW: Can we postpone this discussion?
sandro: I think we're 95% there; we just need to do a few minor things and 
then see whether there are any objections.
ChrisW: Does anyone object?
<PaulaP> I object, the discussion needs to be postponed
<Allen> yes
ChrisW: We will postpone the OWL and SPARQL discussions for a future draft 
of the UCR document.
... Now let's review the 'motivates' links in the use cases.
DavidH: If people haven't reviewed this document yet, they will not 
contribute productively to the discussion.
ChrisW: If you haven't read the use cases section, please reserve your 
comments for later.
... if we can't come to agreement at this time, we can remove the 
'motivate' links from the document (just for this draft)
DavidH: Some of the 'motivate' links are obvious though, and I think we 
should keep those.
ChrisW: But then some use cases will have 'motivates' links and some will 
not, and this will be confusing to the reader.
ChrisW: In the document, the 'Negotiating eBusiness Contracts Across Rule 
Platforms' use case motivates semantic precision (which I think is clearly 
the case), and implementability (which I think is not so clearly the 
case).
DaveR: All of the use cases require semantic precision. Some things are 
just required by RIF, and they either shouldn't be listed or should be 
listed in all use cases.
(discussion about whether there should be links in both directions between 
use cases and requirements)
DavidH: I think it makes more sense to have forward (from use case to 
requirement) links. I don't think that having both forward and backward 
links is necessary.
<PaulaP> yes
DavidH: Regarding use case 2.1: Peter PS suggested that 'compliance model' 
and 'coverage' are not necessary.
ChrisW: and what about 'Implementability'?
Paula: 'Implementability' refers to implementation of translators
<csma> no
<csma> coverage
PaulaP: Yes, I agree that maybe this use case (2.1) does not motivate 
implementability.
<Francois> sorry, but I must leave now.
PaulaP: The authors of the use cases are most familiar with the use cases, 
so perhaps they are the best suited to add in the correct links to 
requirements.
DavidH: This is true, but then since different people would be updating 
different use cases, we may get inconsistency.
PaulaP: In that case, perhaps it is better to leave out all 'motivates' 
links from use cases to requirements for now.
Christian: The 'coverage' requirement is motivated by all the use cases. 
Therefore, if we leave the 'motivates' links in this version of the UCR 
draft, then we should update it so that the 'coverage' requirement is 
motivated by all use cases.
<PaulaP> +1 to csma
ChrisW: 'Semantic Precision' and 'Coverage' are two requirements that are 
also motivated by all use cases.
DavidH: Well, there are degrees of motivation. Some use cases motivate 
semantic precision more than others.
AllanG: I think it makes sense to either leave the 'motivates' links out 
completely for the time being, or else change them to something stronger, 
i.e. to indicate that the use case can't do without it. Perhaps 'require' 
instead of 'motivate'. But this may be too big of a change to do right 
now.
DavidH: If we don't have motivation links, then the reasons that we have 
the new use cases will not be clear
Christian: How about if we change 'motivate' to 'uniquely motivate'?
ChrisW: Does use case 2.1 uniquely motivate any requirement?
PaulaP: In some use cases, the form of the data is not mentioned, so I 
don't think it uniquely requires any specific format
AllenG: 2.1 doesn't 'uniquely require' (?'XML data'?), but it does require 
it as opposed to just motivating it.
ChrisW: It will be a substantive change to put the 'motivates' links in 
order. Given the current deadline, it is perhaps better to remove the 
'motivates' link from this draft. ... The links are important though; we 
can add them in a future draft.
<PaulaP> +1 to remove the links use cases - requirements
<csma> +1 to remove the motivate links for now
DavidH: What does Sandro think about this? I think we are close to having 
it figured out?
Sandro: I agree with Chris. We don't have time to figure it out now.
<scribe> ACTION: David H: Remove the motivates links. [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action04]
ChrisW: DavidH, we need to make sure that the 'motivates' links topic does 
not get dropped for future versions. Please keep an eye on that.
PaulaP: Use case 2.8 describes its rules in a different structure from the 
other use cases.
DavidH: I could change the representation of the the rules for use case 
2.8 to be more like the others. However, this use case deals with mapping 
data, and that is the reason that its rules are described differently.
DaveR: Yes, it is easier to map data structures in a semi-formal way, 
rather than with natural language. I think it's ok as is: it isn't in any 
specific formal rules syntax.
<DavidHirtle> we're talking about the rules here:
<DavidHirtle> 
http://www.w3.org/2005/rules/wg/ucr/draft-20060626#Vocabulary_Mapping_for_Data_Integration
DaveR: Use Case 2.7 also uses a semi-formal syntax to describe the rule, 
so if we want consistent representation style across all rules in the use 
cases, then we'll have to change that one too, in addition to 2.8.
ChrisW: Who objected to use case 2.8?
PaulaP: I don't object to leave them as they are; it was just a question.
Sandro: Peter PS made the point that we should have a disclaimer regarding 
rules controlling human behavior
<sandro> Peter's Comments, ChrisW: 
http://lists.w3.org/Archives/Public/public-rif-wg/2006Jun/0149.html
ChrisW: The introduction already mentions an intentional lack of syntax. 
Maybe we should just add a small disclaimer along the lines of what 
PeterPS suggested.
<ChrisW> However, this informality can lead readers to the conclusion that
<ChrisW> rules can perform arbitrary actions in the real world. This is 
not
<ChrisW> the case - the RIF WG has not yet decided on the ultimate power
<ChrisW> that rules will have.
ChrisW: Does anyone object to adding PeterPS's suggested further 
disclaimer?
DavidH: In addition, I think we should add this phrase to the end of the 
disclaimer: "except where doing so would detract from readability."
<sandro> +1 "Except where doing so would detract from readability"
<sandro> +1 Peter's further disclaiminer
<sandro> RESOLVED
ChrisW: There are no objections to updating the introduction to the UCR 
document with the additional disclaimer suggested by PeterPS and 
additional clarification suggested by DavidH.
<scribe> ACTION: David H to update use cases introduction [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action05]
<scribe> ACTION: csma to ask Frank to merge first two CSF's [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action06]
<scribe> ACTION: Paula P to update the text to merge the first 2 CSF's 
[recorded in http://www.w3.org/2006/06/27-rif-minutes.html#action07]
ChrisW: Frank McCabe will update the reqs/goals/csf diagram to show the 
merged alignment CSF. Paula P will make the corresponding update to the 
associated text.
<PaulaP> +1 to adding an overview
<scribe> ACTION: Allen G to add an organizational overview to the 
introduction by 6/27 [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action08]
ChrisW: Everyone who has an action should send an email to Sandro when 
they are done so that he can make the revised draft of the UCR document.
ChrisW: We planned to talk about a process for populating RIFRAF today, 
but we are out of time. This will be discussed later. 
ChrisW: Any other business to discuss today? ... None. Adjourned.
<DavidHirtle> sandro, can we talk for a moment after?
Summary of Action Items
[NEW] ACTION: Allen G to add an organizational overview to the 
introduction by 6/27 [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action08]
[NEW] ACTION: ChrisW will check that releasing PRR draft to RIF members is 
ok with IBM [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action01]
[NEW] ACTION: csma to ask Frank to merge first two CSF's [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action06]
[NEW] ACTION: David H to update use cases introduction [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action05]
[NEW] ACTION: David H will add the XML req to ? [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action03]
[NEW] ACTION: David H: Remove the motivates links. [recorded in 
http://www.w3.org/2006/06/27-rif-minutes.html#action04]
[NEW] ACTION: Leora will add text to introduction - will change it in the 
wiki [recorded in http://www.w3.org/2006/06/27-rif-minutes.html#action02]
[NEW] ACTION: Paula P to update the text to merge the first 2 CSF's 
[recorded in http://www.w3.org/2006/06/27-rif-minutes.html#action07]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.127 (CVS log)
$Date: 2006/06/27 16:32:10 $