RIF Telecon 23 Jan 07 -- 23 Jan 2007

Admin

csma: DaveR is the editor for the glossary

<ChrisW> ACTION-203 closed

<ChrisW> ACTION-204 closed

<ChrisW> ACTION-208 closed

<ChrisW> ACTION-209 closed

<csma> http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0108.html

resolved: 1/9 telecon minutes approved

<csma> http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/att-0101/16-rif-minutes.html

1/16 telecon minutes not yet approved

F2F

<sandro> http://www.w3.org/2002/09/wbs/38457/f2f5reg/

<Allen> done

csma: only one hotel in area for the f2f?

allen: there are more than one, but this one is least expensive and has shuttle service

<ChrisW> sandro, is the f2f reg form up?

allen: could put other hotels on page, but can't guarantee shuttle

<ChrisW> ACTION: Allen to add some hotels to f2f5 page [recorded in http://www.w3.org/2007/01/23-rif-minutes.html#action01]

<rifbot> Created ACTION-210 - Add some hotels to f2f5 page [on Allen Ginsberg - due 2007-01-30].

allen: please everyone fill out the f2f questionnaire -- we need the information at least 5 days before

sandro: current form closes on feb 10

Liaison

<johnhall> sbvr - no change

csma: nothing to report

UCR

<ChrisW> ACTION 144 done

csma: added chapter called analysis in UCR document, added section on usage scenarios
... (didn't know where else to put the section)
... can we just drop issue 12 without any specific action?

<josb> +1 to Chrisw's statement

daveR: I just give in at this stage

csma: any objections to dropping issue 12?

daveR: by saying nothing, we're not precluding possibility of doing the work along the way

csma: didn't resolve either way -- don't write anything in the document yet

<sandro> ChrisW: The fact is, proposals in each direction have failed. We haven't agreed to do it or to NOT do it.

chrisW: right

<sandro> +1 drop issue 12

csma: proposed: drop issue 12

resolved: drop issue 12

csma: are we ready to release UCR WD3?
... (after review etc.)

csma: how long to prepare final version so we can have internal review?

allen: probably by next week

chrisW: document hasn't changed fundamentally, just new pointers etc. so I'd be happy with short review period

sandro: try it and see if anyone objects

csma: agreed, 1 week review period starting next week, with vote two weeks from now

RIFRAF

axel: something doubled in current draft, coverage and RIFRAF
... trying to pick RIFRAF back up

csma: we decided RIFRAF was not on critical path for now (neither UCR nor technical design)

axel: what's in there isn't last version of RIFRAF wikipage
... e.g. discriminators missing

csma: is it feasible to edit updated version by next week?

axel: if we just copy and paste, no problems, but there have been other suggestions

<LeoraMorgenstern> I believe they should be discussed afterward

<LeoraMorgenstern> I just sent email about this:

leora: there's lots of work left to do on the discriminators -- not that they're wrong, but somewhat arbitrary/incomplete
... need more comprehensive set, but RIFRAF as it stands is sparse

<Hassan> +1 with Leora's statement on completeing and review the discriminators

<Harold> As I mentioned, there are further possible discriminators, e.g. 'open' vs. 'closed' slots.

csma: the next big step in UC&R should be RIFRAF, but for this time we can leave as is

<PaulV> +1 on Leora's comments too: discriminators should be wrt to interchange issues only?

leora: I think we should start back up on RIFRAF because at some point will be on critical path

<Zakim> sandro, you wanted to advocate ontologizing what we need

sandro: we could fill in skeleton of what we've discussed

<AxelPolleres> +1 5min more , no more

leora: I think it's a little dangerous to put in only what we've discussed so far

<AxelPolleres> sent three mails on how to proceed with RIFRAF yesterday: http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0104 http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0105 http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0106

leora: we need better way to decide what to include versus not (not just special interests of people)

sandro: why do we need that coverage?

<sandro> Sandro: knowing why we need the descriminators will help us understand how to develop RIFRAF

axel: I started trying to get something out of the core (as in my email)
... should we set a date (e.g. 3 weeks) for other people to try to ontologize part of the questionnaire?

csma: need first ontology before deciding what to include versus what not

<AxelPolleres> http://lists.w3.org/Archives/Public/public-rif-wg/2007Jan/0105 contains a core proposal for the ontology.

csma: if some of the people on the RIFRAF list can commit to doing something in next 3 weeks, we can start to discuss in 3 weeks

<sandro> I don't think I can commit to working on RIFRAF in this time-frame (3 weeks)

<AxelPolleres> the assignments were made in the last f2f.

<PaulaP> I don't have time for RIFRAF during the next 3 weeks

csma: would it be possible to have first version in time for f2f?

<ChrisW> axel, will you be at the f2f?

<PaulaP> and I won't be at the next F2F

<AxelPolleres> ChrisW: sorry, I won't be there physically

<ChrisW> Leora?

<PaulaP> +1 for discussing RIFRAF after F2F

csma: some won't be at f2f and don't have time before, so let's discuss RIFRAF after f2f
... can start working before, of course

leora: moving discussion until after is fine by me
... main thing is that we don't seem to have methodology for discriminators

<AxelPolleres> I suggest the involved people to follow up my mail.

<LeoraMorgenstern> Axel, I did respond

csma: let's move methodology discussion to email -- Leora can start it

<AxelPolleres> Leora, I what think we should follow up can be done if you all look at the OWL doc I sent around and make suggestions where/how to extend it, which we can do by mail.

Tech Design

sandro: hard at work on action 182

csma: consensus on leaving constraints out of design WD1

<Harold> Michael is calling through gizmo.

<sandro> ugh -- sound quality is lousy, MichaelKifer --- sorry for recommending gizmo !

csma: did you say we should have at least some primitive types, Michael?

michaelKifer: yes

<AxelPolleres> XSD simple types and XPath/XQuery functions may be starting points.

<Harold> Date, Time, ...

<ChrisW> the sound sof the keyboard are clear at least

<MichaelKifer> we can include dateTime, integer, string, etc.

<agiurca> Alex +1

axel: these are starting points for datatypes and builtins

csma: the charter says at least some types should be in core, right?

sandro: could be lots of work -- a lot to ask everyone to handle all date operations

<ChrisW> In Phase 1, the format must support literals and common functions and operators for at least: text strings (xsd:string), 32-bit signed integers (xsd:int), unlimited-size decimal numbers (xsd:decimal), Boolean values xsd:boolean), and list structures.

csma: who would be against having date & time types in core?

chrisW: I think we should postpone to get WD1 out

<MichaelKifer> The point is that without concrete exampoles of types there will be difficulties in understanding the framework

<GaryHallmark> durations are painful

<MichaelKifer> is duration different from time?

<DavidHirtle> yes

<Harold> Michael, could we also illustrate multisorted logic with user-defined types?

michael: I don't see a problem with having dateTime in WD1

daveR: of those in charter, I would recommend leaving out list

<Hassan> ints, floats, strings, chars, lists, fots

<GaryHallmark> timezones can muck up datetime

<MichaelKifer> we dont need lists - they are there if we have function symbols

hassan: lists are just one kind of first-order term

daveR: I only meant in terms of XSD datatype

<Harold> Currently, http://www.w3.org/2005/rules/wg/wiki/CORE/Conditions/Positive mentions "sorts integer, strings, time, dates, etc.", so we could keep this as is.

csma: proposed list: int, string, time, date

<MichaelKifer> real also

hassan: what about decimal and floats?

<sandro> decimal are exact -- used for money.

<Harold> Many other standard bodies are working on this...

<Harold> Can we avoid going into this now?

<Harold> IEEE, ...

<AxelPolleres> what do you mean by reals, but reals can't be represented properly in machine readable form..., can they?

<MichaelKifer> ok

csma: agree so far that in first core WD we have no constraints, no slots, no built-ins and have a few primitive types (i.e. strings, integers, reals, decimals and time date)

<AxelPolleres> +1 to decimals, -1 to reals

<FrankMcCabe> got to go guys ...

<josb> decimals as they are in XML schema, correspond to the usual real numbers in pure mathematics

hassan: reals are not real datatypes
... reals are just mathematical concept, while floats are IEEE standard

csma: do we really need both floats and decimals in first WD?

<AxelPolleres> floats are represented by exponent and mantisse, reals are something like e.g. sqrt(2), periodic numbers, which can't be properly represented finitely.

<ChrisW> PROPOSED: 1st CORE WD will not include constraints, slots, builtins

csma: proposed resolution: leave out slots, constraints and built-ins in core WD1

<josb> reals can indeed not always be represented finitely

sandro: I don't understand -- I thought data types and built-ins are linked

<PaulV> PaulV has to depart

<PaulV> bye

<josb> with floats you can have this kind of strange thing: 1+1-1=1,0000001

<sandro> MichaelKifer: yes -- we'll leave out the builtins in the first WD, but we'll have a note which says the builtins will come later.

<sandro> sandro: fine with me

<ChrisW> RESOLVED: 1st CORE WD will not include constraints, slots, builtins

csma: first WD of core will include primitive types: string, integer, decimal and dateTime

(no objections)

<ChrisW> RESOLVED: 1st CORE WD will include the following datatypes: string, int, decimal, datetime

<Harold> We should also keep the initial XML syntax, which has been in the CORE from the beginning and has been already employed to illustrate use cases: Rulebase (or Ruleset), Implies, And, Atom, Expr, Con, and Var.

harold: should keep initial XML syntax to make it more clear that we're doing an XML-based syntax

<AxelPolleres> (just to be complete: periodic numbers can be represented finitely buy rational numbers ;-) remembering from school maths)

<agiurca> XPath are builtins

csma: couldn't it be confusing if we have two XML syntaxes in core?

harold: need to see both the source and target of transformation

csma: my concern is people could focus on this XML syntax and start playing with it, missing the whole point

chrisW: I don't think we've had enough attention paid to that particular XML syntax

<AxelPolleres> It would be strange if the final goal wasn't to have an XML syntax, don't we even have this in the requirements/charter?!?

csma: at minimum we should mention that there's going to be an XML syntax

<Harold> Actually, the XML syntax is very close to the abstract syntax (in the sense of John McCarthy).

<AxelPolleres> If we have one, why not putting it out there for comments? We can still change in the next draft...

<AlexKozlenkov> We do need XML syntax to talk to people. For example, I'll need it to talk to our business people

hassan: I think Harold just wants to use his tags for illustration, i.e. say in draft of this kind (not necessarily this one)

<agiurca> Examples can be done also in the abstract syntax

<Harold> The human-oriented syntax is actually a concrete (ASCII) syntax.

<sandro> Can we come back to this point next week, when I have my action done?

<ChrisW> we definitely NEED an xml syntax - that's not the point

<DavidHirtle> +1 to Axel

hassan: examples are good

<AlexKozlenkov> The point is we need it fast before the interest starts to fade away

sandro: can have my action done Friday

<Harold> Sandro, deriving means one needs a source and a target. What is your target?

<AxelPolleres> Chris: Just a last question on that. If we have something, and we could include it, aren't even comments on the tags from outside valid as input?

csma: could discuss this next telecon

axel: if we have one, why not put it out for comment?

csma: my point is this is something we didn't discuss this -- comments can be useful, but don't want people to focus on something we didn't even discuss

<ChrisW> the xml syntax was presented as an exmaple of what an xml syntax might look like

csma: let's discuss this next meeting

<ChrisW> it was never scrutinized to the degree we have looked at the semantics, e.g.

<AxelPolleres> got the point.

csma: harold, do you feel comfortable with drafting a section about extensibility dialects and conformance?
... my understanding is that we'll release WD1 before anything like the arc document

harold: yes, will do

csma: w.r.t. working with OWL and RDF, possible to have something in WD1?
... okay to just mention this is work in progress?

chrisW: my sense is that it needs more discussion

harold: we are chartered to have separate document about those

chrisW: not necessarily separate

<Harold> We did have: http://www.w3.org/2005/rules/wg/wiki/A.5_Extension%3A_Ontology_Conditions

daveR: I think that section needs more discussion

<ChrisW> +1 to DaveR

csma: let's leave that open for now -- we'll see closer to WD1

<Hassan> +1 with David Reynolds as well

csma: would like to have a reasonable mention of OWL and RDF

<Harold> Can we update http://www.w3.org/2005/rules/wg/wiki/OWL_Compatibility and http://www.w3.org/2005/rules/wg/wiki/RDF_Compatibility ?

<Harold> ... and refer to those in a way?

<ChrisW> i think it's ok as long as the reference is clear that this is a "possible approach" we are considering

csma: when can you reasonable expect a complete first draft?

harold: perhaps by mid-February

chrisW: Feb 13th?

harold: yes

csma: Feb 9th (Friday before telecon)

harold: yes

<ChrisW> Christian is tougher than me

AOB

csma: adjourn

<ChrisW> +1

<PaulaP> +1

<Hassan> +1

<PaulaP> bye

<agiurca> bye

- DRAFT -

RIF Telecon 23 Jan 07

23 Jan 2007

Attendees

Contents

Admin

F2F

Liaison

UCR

RIFRAF

Tech Design

AOB

Summary of Action Items

Scribe.perl diagnostic output