RIF Telcon 1-Jul-2008 -- 01 Jul 2008

<csma> list agenda

<csma> Axel, would you like to scribe?

:-)

<scribe> scribe: Axel Polleres

<scribe> scribenick: AxelPolleres

<ChrisW> Harold, are you joining?

csma: Any agenda changes?

<csma> http://lists.w3.org/Archives/Public/public-rif-wg/2008Jun/att-0172/rif-minutes-24062008.html

csma: no

<ChrisW> PROPOSED: Accept last week's minutes

csma: objections to accept the minutes of last week?

<ChrisW> RESOLVED: Accept last week's minutes

csma: none. accepted.

ChrisW: were minutes two weeks ago accepted... yes.

<csma> next item

Liaison

csma: Wehad an OMG PRR meeting last week, deadline for final report has been moved 3 months from now.
... we try to delay PRR to "wait" for RIF PRD.

<csma> next item

F2F11

csma: date is fixed.

<sandro> Sep 26-27 (Friday/Saturday)

ChrisW: Yes, it is: Sep 26-27.

csma: location: Manhattan or Hawthorne.

<josb> +1 Manhattan

+/-0

csma and sandro discussing hotel prices in NY.

ChrisW: Hotels considerably more xepensive, also transport.

<sandro> http://www.w3.org/2005/rules/wiki/F2F11

Action Review

<scribe> ACTION: 534 to CONTINUED. [recorded in http://www.w3.org/2008/07/01-rif-minutes.html#action01]

<trackbot> Sorry, couldn't find user - 534

Action-530: CONTIINUED.

<trackbot> ACTION-530 Review final BLD LC draft notes added

<ChrisW> Harold, are you joining?

Leora: The resolutions in the documents are not always clear.

<MichaelKifer> harold will be late

Leora: e.g. http://lists.w3.org/Archives/Public/public-rif-wg/2008Jun/0166.html

csma: references to DTB have been clarified by last weeks resolution.

Leora: When will BLD be frozen enough to complete the action?

<josb> the RDF namespace is already used for XMLLiteral

<csma> http://www.w3.org/2007/OWL/wiki/InternationalizedStringSpec

<sandro> http://www.w3.org/2007/OWL/wiki/InternationalizedString

ACTION-521: CONTINUED.

<trackbot> ACTION-521 Work with Jie Bao (of OWL-WG) to put together draft on rif:text/owl:internationalizedText notes added

<ChrisW> ACTION: sandro to work on getting access to OWL wiki for internationalized text article [recorded in http://www.w3.org/2008/07/01-rif-minutes.html#action02]

<trackbot> Created ACTION-535 - Work on getting access to OWL wiki for internationalized text article [on Sandro Hawke - due 2008-07-08].

josb: sandro, question about the previous item. two links on the OWL side on that side, what is the difference?

ACTION-517: continued.

<trackbot> ACTION-517 Review FLD [june 23] notes added

ACTION-517: all other review actions also continued.

ACTION-510: continued.

<trackbot> ACTION-510 Remove UPDATE, EXECUTE and ASSIGN from PRD notes added

sandro: no point to talk about actions on metadata without harold and michael.

Publication plan.

BLD

Michael: We are still working on some aspects of PS to XML translation.
... made some signnificant improvements, not yet finished.

csma: you can use some example from PRD, maybe.

Michael: we define a recursive function. but still have left the translation of Prefix, Base, will take only a couple of days more.
... should be doable in one or two days, depends on Harold's schedule.

josb: looked over the XML translation... looks good. but I thought we don't need to define translations for Base and Prefix.

Michael: well, they are part of the language, so we should explain it.
... I don't see how we get around this.
... we have to define the translation to entity definitions.

Harold: (just joined) looked it up, it can be done.

csma: our question was, when would we have it finished?

Harold: two days

csma: so we can freeze BLD on Friday.

Harold: yes.

josb: we need two things entities and xml-namespaces for Prefix.

Harold: ... no it is all only between the tags, so we don't need namespaces.

<Harold> For pretty print, switch to XML: http://www.w3.org/TR/2008/WD-owl2-primer-20080411/#Basic_Notions

<Harold> Then view: http://www.w3.org/TR/2008/WD-owl2-primer-20080411/#Appendix:_The_Complete_Example

<Harold> Harold

Michael: They do XML namespaces in OWL/XML, but it doesn't seem necessary.

Harold: I have pasted some links that illustrate this.
... the namespace definitions don't seem to be necessary in those examples.

ChrisW: we seem to all agree here.

josb: why do we need entities/namespaces at all? we can just expand at preprocessing.
... but I am not very concerned with it.

Michael: We can state that all prefixes should be preprocessed. Don't know whether this is simpler, though.

I tend more towards describing the preprocessing of prefixes.

josb: would prefer preprocessing, but don't oppose other option (entity definitions)

Harold: preprocessing can be implemented in the document quickly.
... we also need to clarify the MIME type question.

sandro: We need a MIME type registration.

harold: I just replaced RDF by RIF in some template from RDF for that.

sandro: when we go to LC, we have to have the MIME type registered/acknowledged.

csma: who can finish that?

sandro: none of us has experience with that.

<josb> http://www.w3.org/2001/sw/RDFCore/mediatype-registration

<Harold> Michael and all, Ys we would need to add a sentence or two saying something: The presentation syntax for Base and Prefix will be preprocessed by expanding the Base and Prefix names into full URIs. Only that expanded version will then be translated to XML as shown in this section.

josb: do we need some formal/informal registration with IANA?

<ChrisW> Harold - Just put that in the function

josb: RDF had it in a separate document.

sandro: doesn't need to be a separate doc, can be an appendix in BLD.

Harold: need some help/feedback on the XML Schema.

<ChrisW> ACTION: sandro to finish mime type for BLD [recorded in http://www.w3.org/2008/07/01-rif-minutes.html#action03]

<trackbot> Created ACTION-536 - Finish mime type for BLD [on Sandro Hawke - due 2008-07-08].

csma: who will finish the mime type stuff?

sandro: will do it/clarify it by friday.

<Harold> RE: Media-Type (MIME type) for RIF http://lists.w3.org/Archives/Public/public-rif-wg/2008Jun/0070.html

Harold, josb, axel: does not affect SWC and DTB.

josb: can do a review on Wed July 9th.

<csma> PROPOSED: Mark "external frames" AT RISK in BLD

<Harold> RE: Media-Type (MIME type) for RIF http://lists.w3.org/Archives/Public/public-rif-wg/2008Jun/0072.html

csma: we want to mark external frames at risk in bld since we don't understand what they are.

<ChrisW> ACTION: harold to mark external frames "at risk" in BLD [recorded in http://www.w3.org/2008/07/01-rif-minutes.html#action04]

<trackbot> Created ACTION-537 - Mark external frames \"at risk\" in BLD [on Harold Boley - due 2008-07-08].

<csma> RESOLVED: Mark "external frames" at risk in BLD

csma: when can you do the review, leora?

leora: over the weekend.

ChrisW: the whole WG is responsible to take a look as well!!! i.e. if no more complaints, jos and leora's reviews will be the last step before going the next step to LC, after jos and leora's reviews.

csma: this means, jos and leora, make clear in your review whether it is a "Go", or a "No Go"

SWC, DTB

ChrisW: same question now to jos: when do you expect to freeze SWC?
... and same for DTB, Axel?

josb: after dtb.

axel: DTB is frozen from my side.
... (for first WD), but would like to have another review, of course.

josb: SWC will be frozen by Tue, COB.

csma: I will review SWC, can give go-nogo, Tue, 15th, COB.
... Michael, when is FLD frozen?

FLD

Harold: only XSD missing... will be one week form now.

Michael: we have to fix the table and force other improvements from BLD, but can be done within a week.
... after next telecon, Friday 11th.

csma: who can review?

chrisW, josb: by 15th.

<ChrisW> zakkim, list agenda

csma: now what about PRD, ChrisW, pls take over chairing.

PRD

ChrisW: discussions on the mailinglist.
... Christian, you sent a list of interim solutions for 1st WD compromises.

Gary: I had a look, sounds great.

csma: I had some things worked out already, few small edits still to implement.

<markproctor> for no-loop yes.

<markproctor> ORB is just built on jess.

Gary: (would you might typing in your remark yourself)

ChrisW: Let us discuss 1st WD schedule and then get back to details.

<markproctor> end of the day it comes down to. If you change a fact and for the same rule and same matched data it re-matches do we activate or not.

<markproctor> by default jrules does not.

<markproctor> jess/drools/clips will re-match unless no-loop is used.

csma+Gary discussing some resolution on conditional retract/assert

csma: will have all changes implemented tomorrow.
... adrian and gary are the reviewers.

chrisW: Can you take an action to give the final go/nogo for 1st WD?

<ChrisW> ACTION: Gary to review frozen PRD WD Thursday [recorded in http://www.w3.org/2008/07/01-rif-minutes.html#action05]

<trackbot> Created ACTION-538 - Review frozen PRD WD Thursday [on Gary Hallmark - due 2008-07-08].

Gary: will be gone 4-20th, but can review on Thu.

ChrisW: We will have PRD frozen tomorrow and Gary giving final go/nogo for WD on Thu.
... any more technical discussions necessary?

<ChrisW> discussion of "norepeat"

<markproctor> ilog does 'no-repeat" as default.

<markproctor> drools, jess, clips do not - you must specify the no-loop attribute for the rule.

<markproctor> no you can't do that.

<markproctor> you cant do it per rule

<markproctor> it has to be per rule + row of data.

csma: semantics of norepeat is explained exactly like Gary wants in the current proposal.

<markproctor> no rule execution, except of the data, is not really a valid use case.

<GaryHallmark> http://www.ilog.com/products/jrules/documentation/jrules67/rssamples/rs_smp_rsruleprog2.html

<GaryHallmark> "The default refraction mechanism avoids executing the same rule several times on the same object instances, even if the rule conditions are met again after a change on an object."

<markproctor> hwo do I get on the phone queuue?

Gary: that doesn't seem to be what ILOG does in their implementation.

<markproctor> correct - the row of data is the important bit.

csma: will check back.

<markproctor> guess no more needs to be said now.

<markproctor> AxelPolleres: i've never actually run JRules, so can't qualify. can just say what I've been told and read, and drawn from my own understanding of PR systems.

csma: am sure to be finished by tomorrow, apart from norepeat, may be marked in the draft.

Gary: Should we have architectural principles to minimize divergence for new dialects from BLD/Core?
... I'd like to have feedback from the non-PR people here.

ChrisW: hard to follow for the non-PR people who didn't read the whole thread in detail.

csma: I'd urge all people to give their opinions on the PRD draft.

<markproctor> did everyone get my excel file I emailed to the list?

csma: my view is slightly different from Gary's. My concern is more putting in balance to make the mainstream PR engine people happy over maximizing BLD overlapping.

<markproctor> does it help at all? maybe we can grow it further to correct different engine capabilities.

<markproctor> will help decide on features to support.

<markproctor> just added no-loop to my list.

Gary: I have a hard time giving a handle on the argument here, Clips looks pretty much like BLD, for instance.

<markproctor> is there a website I can put this spreadsheet?

<markproctor> or some google docs?

<josb> use the wiki

<csma> Mark, I did not have an opportunity to look at it yet. But that's certainly useful.

<markproctor> csma: I apologise if anything is wrong.

ChrisW: I don't think that following PR engine people is the most relevant, let them speak for themselves within the group.

<markproctor> guys you are adding language semantics now, can we move on?

<markproctor> sorry s/adding/argueing/

csma: I think mark's work in his excel sheet is interesting.
... I am arguing for publishing PRD as early as possible.

<markproctor> no it's fine.

<Harold> Chairs, we need to come back in the ongoing telecon to freezing BLD for a moment: As has been often discussed, Equal roles need to be left and right: http://lists.w3.org/Archives/Public/public-rif-wg/2008Jun/0099.html

<markproctor> ok, i'm trying to make up some wiki table now for htis

AOB (getting back to equality in BLD/XML)

ChrisW: do we have some time to talk about casting?

Harold: we need to split the role of LHS and RHS in equals.

<csma> PROPOSED: Change the roles in Equal from "side" to "left" and "right" (in BLD XML)

Topic went back to BLD/XML ...

<josb> -1

<Harold> You cannot even say Equal(x y) = Equal(y x) without splitting side into left and right.

<Hassan> Such is called a rewrite rule, Harold. So why not call it that instead of equation?

<Harold> This is the job of the semantics.

Harold: reasons: we cannot talk about symmetry of equality in the semantics, if they are symmetric per definiton in the syntax.

<sandro> jos: I would not oppose changing names in *XML*.

josb: it should be interpreted as identity.
... I don't care about BLD/XML, though.

<Hassan> I favor lhs/rhs rather than side/side (actually I did from the start) ...

Michael: in PS there is a distinction between left and right of the '=' sign anyway.

josb: then fine, if it is only in the XML syntax.

<markproctor> how do I unlock a page?

<josb> 0 on the resolution

<markproctor> created this - http://www.w3.org/2005/rules/wiki/PR_Feature_comparison

<markproctor> I go to edit and it says it's been locked

<sandro> markproctor, you have to log in, and have a WG-enabled account.

<markproctor> "This page has been locked to prevent editing. "

<csma> PROPOSED: Change the tag name of the sub elements of Equal in BLD XML from side to left and right

<markproctor> ah bugger

-0 (why can't we drop sides tag at all?)

<sandro> "locked" is media-wiki speak for "you don't have write access"

<Hassan> +1

<Harold> +1

<markproctor> I probably had one already, but forget what it is.

<MichaelKifer> +1

<sandro> +1

<ChrisW> 0

<josb> 0

<markproctor> just created a new account, mdproctor

<josb> (don't care)

RESOLUTION: Change the tag name of the sub elements of Equal in BLD XML from side to left and right

<Hassan> +1

<ChrisW> +1

<GaryHallmark> +1

<markproctor> can someone enable mdproctor as a username? and I'll do this wiki page now?

<sandro> markproctor, you didn't have an account before. The general advice would be to use the account "Mark Proctor". spaces, etc, are fine.

<sandro> markproctor, but mdproctor is okay, if you want.....

- DRAFT -

RIF Telcon 1-Jul-2008

01 Jul 2008

Attendees

Contents