14:23:46 RRSAgent has joined #rif 14:23:46 logging to http://www.w3.org/2007/04/10-rif-irc 14:23:54 Zakim has joined #rif 14:24:02 zakim, this will be rif 14:24:02 ok, ChrisW; I see SW_RIF()11:00AM scheduled to start in 36 minutes 14:24:42 Meeting: RIF Telecon 10 Apr 07 14:24:53 Chair: Christian de Sainte-Marie 14:25:17 Agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Apr/0001.html 14:25:37 ChrisW has changed the topic to: 10 April RIF Telecon Agenda: http://lists.w3.org/Archives/Public/public-rif-wg/2007Apr/0001.html 14:25:50 rrsagent, make minutes 14:25:50 I have made the request to generate http://www.w3.org/2007/04/10-rif-minutes.html ChrisW 14:26:08 agenda+ Admin 14:26:15 agenda+ Liason 14:26:23 agenda+ Core 14:26:29 agenda+ UCR 14:26:36 agenda+ RIFRAF 14:26:43 agenda+ AOB 14:26:54 rrsagent, make logs public 14:41:20 Francois has joined #rif 14:47:25 patranja has joined #rif 14:57:16 csma has joined #rif 14:57:30 AlexKozlenkov has joined #rif 14:58:31 Alex, can you scribe today, please? 14:58:35 IgorMozetic has joined #rif 14:58:46 SW_RIF()11:00AM has now started 14:58:54 +??P1 14:59:11 -??P1 14:59:13 SW_RIF()11:00AM has ended 14:59:14 Attendees were 14:59:27 SW_RIF()11:00AM has now started 14:59:33 +??P7 14:59:39 Hassan has joined #rif 14:59:46 zakim, ??P7 is me 14:59:46 +csma; got it 14:59:56 +??P1 15:00:08 +Hassan_Ait-Kaci 15:00:10 ??P1 is me. 15:00:21 zakim, ??P1 is me. 15:00:21 +Francois; got it 15:00:23 +[IBM] 15:00:24 zakim, ibm is temporarily me 15:00:24 +ChrisW; got it 15:00:28 zakim, mute me. 15:00:29 Francois should now be muted 15:00:43 StellaMitchell has joined #rif 15:00:45 sandro has joined #rif 15:01:24 josb has joined #rif 15:01:25 yes 15:01:30 thanx 15:01:49 DaveReynolds has joined #rif 15:01:52 + +39.047.101.aaaa 15:02:15 +[IBM] 15:02:21 zakim, [ibm] is temporarily me 15:02:21 +StellaMitchell; got it 15:02:30 +[IPcaller] 15:02:40 zakim, [IPcaller] is me 15:02:40 +IgorMozetic; got it 15:02:45 zakim, mute me 15:02:45 IgorMozetic should now be muted 15:02:47 scribenick: StellaMitchell 15:02:51 + +49.892.180.aabb 15:03:03 zakim, who is on the phone 15:03:04 I don't understand 'who is on the phone', csma 15:03:09 (I'll be a few minutes late.) 15:03:11 zakim, who is on the phone? 15:03:12 On the phone I see csma, Francois (muted), Hassan_Ait-Kaci, ChrisW, josb (muted), StellaMitchell (muted), IgorMozetic (muted), PaulaP 15:03:22 +??P25 15:03:56 -??P25 15:04:29 zakim, next item 15:04:32 agendum 1. "Admin" taken up [from ChrisW] 15:05:00 csma: next meeting, next Tues, April 17 15:05:01 +Dave_Reynolds (was Guest P28 74394) 15:05:04 +Dave_Reynolds 15:05:36 PROPOSED: approve minutes from March 27 15:05:50 Regrets: AllenGinsberg, AxelPolleres, MichaelSintek, DavidHirtle 15:05:52 RESOLVED: approved minutes from March 27 telecom 15:06:09 csma: any proposed amendments? 15:06:35 ...(none) 15:06:41 zakim, next item 15:06:41 agendum 2. "Liason" taken up [from ChrisW] 15:06:48 csma: action review 15:07:01 action-268 complete 15:07:12 csma: any news from liasons? 15:07:25 csma: no news 15:07:32 zakim, next item 15:07:32 agendum 2 was just opened, StellaMitchell 15:08:01 zakim, next item 15:08:01 agendum 3. "Core" taken up [from ChrisW] 15:08:11 zakim, next item 15:08:11 agendum 3 was just opened, ChrisW 15:08:20 csma: action review 15:08:29 yes 15:08:32 action-271 complete 15:09:23 +[NRCC] 15:09:34 IgorMozetic has joined #rif 15:09:42 action-272 continued 15:09:45 Harold has joined #rif 15:10:12 csma: first item of technical design: issue 30, about rif:uri 15:10:29 csma: there has been email disucssion 15:10:40 csma: has this led to any agreement? 15:10:54 csma: michael kifier is not here 15:11:01 csma: jos, summarize? 15:11:11 Hi, I was late so just joined first on the phone, then on IRC. Someone can tell zakim who I am? 15:11:20 jos: i talked with mk offline and no longer have any disagreement 15:11:48 jos: I thought doc was proposal for a dialect, but it is just basically a framework so it is ok to be more general 15:11:56 +Sandro 15:12:01 q? 15:12:06 csma: but what is rif: uri? 15:12:13 (sorry for being late) 15:12:41 jos: it is a sort, an iri written in normal form 15:12:54 jos: no additional semantic implications 15:13:03 MichaelKifer has joined #rif 15:13:07 jos: dialects may introduce additional restrictions 15:13:15 csma: dave reynolds, any comments? 15:13:33 s/implications/commitments 15:14:00 dave: discussion of whether they should be iri's 15:14:13 dave: wasn't sure about ??? 15:14:28 jos: that would be specified in particular dialects? 15:14:43 csma: jos, do you consider this issue basically resolved? 15:14:58 + +43.512.507.aacc 15:15:03 csma: who would be the one to write the additional text on which everhyone would agree 15:15:12 zakim, who is on the phone? 15:15:12 On the phone I see csma, Francois (muted), Hassan_Ait-Kaci (muted), ChrisW, josb, StellaMitchell (muted), IgorMozetic (muted), PaulaP (muted), Dave_Reynolds, [NRCC], Sandro 15:15:14 jos: not sure that we need additional text 15:15:16 ... (muted), +43.512.507.aacc 15:15:26 zakim, aacc is me 15:15:26 +MichaelKifer; got it 15:15:34 csma: the issue is a question re: what is defn of rif:uri 15:15:55 csma: so we need a written definition 15:15:58 s 15:16:16 s/defn/definition 15:16:25 zakim, mute me 15:16:25 MichaelKifer should now be muted 15:16:49 zakim, unmute me 15:16:49 MichaelKifer should no longer be muted 15:17:12 csma: mk, who would write the text that will allow us to close issue-30 15:17:37 csma: WD1 currently refers to the issue 15:17:50 mk: what is the problem with the current text? 15:18:05 jos: it refers to an open issue 15:18:35 mk: is the syntax the same as rfc ? 15:18:43 http://www.rfc-editor.org/rfc/rfc3986.txt 15:19:00 RFC 3987 is the IRI RFC 15:19:10 csma: is it ok to not have any agreed upon definition of rif: uri 15:19:19 http://www.rfc-editor.org/rfc/rfc3987.txt 15:19:26 q+ 15:19:32 mk: there is a definition, but we are not sure it's what we want 15:19:33 ack me 15:19:37 ack dave 15:20:03 daver: on question of syntax, I did propose a phrasing 15:20:21 mk: i think you proposed that the rfc allows relative uri's 15:20:38 mk: and relative uris is just a shorthand 15:21:14 daver: normalizations and change references to iris 15:21:22 An XML syntax for URIs/IRIs was discussed before: http://www.w3.org/2005/rules/wg/wiki/A.1.1_Basis%3A_Positive_Conditions_over_Bipartitioned_Constants 15:21:42 http://lists.w3.org/Archives/Public/public-rif-wg/2007Mar/0133.html 15:21:44 csma: proposed resolution of issue-30 is that we remove sentence from WD! 15:21:51 GaryHallmark has joined #rif 15:21:51 q+ 15:21:54 q+ 15:21:54 ...about open issue 15:21:58 ack harold 15:22:03 ...and inclused dave's proposed text 15:22:16 15:22:24 jos: in earlier doc we already had synax for iri, are people still aware of it? 15:22:32 +Gary_Hallmark 15:22:38 jos: it was an attribute 15:22:49 ^jos^harold^ 15:22:53 Now: 15:23:01 ack jos 15:23:06 s/WD!/WD1/ 15:23:26 q+ 15:23:43 ack dave 15:23:44 jos: for RDF group was not sufficient to just reference rfc, so it might not be in our cas either 15:23:52 csma: that is about syntax 15:24:05 dave: reason for that is iri spec wasn't finished at that time 15:24:20 dave: difference ended up being in treatment of spaces 15:24:33 3987=IRI 15:24:46 dave: sparql was fine pointing to rfc 3987 15:25:12 mk: will we call them uris or iris. have we decided on iris? 15:25:40 harold: most people prefer iri; but i like uri or subset of iri 15:25:57 csma: any objection to using iri? 15:26:12 RDF, SPARQL use IRIs 15:26:15 ack sandro 15:26:18 chris: I don't understand consquences of the decision to use iri 15:26:19 q+ 15:26:28 chris: I don't know of anyone who actually uses them 15:26:37 sandro: concerned about market perception 15:26:48 +1 to Sandro; better to use IRIs, but call them URIs 15:26:56 +1 to holding off on IRI vs. URI 15:27:02 sandro: most people still use "uri" 15:27:19 sandro: uri and iri are mappable to each other 15:27:24 q- 15:27:33 sandro: so in that sense they are equivalent 15:27:50 q+ 15:27:54 csma: i agree with chris, so would rather be more conservative and use uri 15:28:07 csma: let's raise an issue about whether to use uri or iri 15:28:28 ack dave 15:28:34 mk: but dave's formulation refers to iri 15:28:53 dave: people do use iri in rdf data 15:29:39 dave: so if we communicate that rif doesn't support iri, it would surprose people 15:29:59 mk: in rdf, do people use uri or iri? 15:30:11 dave: (didn't catch answer) 15:30:25 chris: how about if the text says "iri or uri" 15:30:36 I wonder if the URI/IRI issue wouldn't be a (much) broader topic than for RIF or even the Semantic Web, perhaps for the Technical Architecture Group (TAG): http://www.w3.org/2001/tag/ (BTW, how do we liaise with the TAG?)? 15:30:51 chris: the name of the sort can still be "rif:uri" 15:31:07 mk: how about use uri now, and say that in future rif might change to iri 15:31:12 csma: still an issue 15:31:27 sandro: i agree with dave that semantic web community expects iri 15:31:53 sandro: people are use to iris, they just don't call them iris 15:32:10 s/use/used/ 15:32:13 q+ 15:32:37 sandro: we agree there is an issue. question is what does the draft say until the issue is resolved 15:33:05 using IRI would be playing safe (wrt. public perception) 15:33:10 ack harold 15:33:16 -IgorMozetic 15:33:41 -1 to Harold 15:33:53 Josb, using URI would be playing safe (wrt technical implementation), as I understand 15:34:08 harold: starting with uris is safer because it is easier to express uri as iri, then iri as uri 15:34:18 any IRI can be encoded as a URI 15:34:18 s/then/than/ 15:34:33 http://lists.w3.org/Archives/Public/public-rif-wg/2007Mar/0133.html 15:34:33 csma: we need reference to dave's text 15:34:35 I propose: Symbols of this sort have the form "XYZ"^^rif:uri, where XYZ is a URI as specified in RFC 3986 (or an IRI as specified in RFC 3987) 15:35:00 chris: i put some proposed text in the irc 15:35:19 chris: I think everyone is saying we will have uris, and we should also include iris 15:35:32 sandro: i think rdf uses iri, but doesn't call them that 15:35:50 chris: rdf in spec, doesn't use iri - it defines a mapping 15:36:09 sandro: that encoding is the same as iri 15:36:18 sandro: is there a normalization step? 15:36:48 csma: chris, in your wording, what is the semantic of the "or" 15:37:04 chris: yes, we will still raise an issue 15:37:11 +1 Chris's wording 15:37:55 PROPOSED: remove from WD1 the sentence about open issue, add dave's proposed text from link above 15:38:15 ... but replace one sentence with chris' text, and close the issue 15:38:39 chris: let's see the changes and then close the issue next week 15:38:47 dave: in what sense close the issue? 15:39:08 dave: so it gets replace with another issue about the syntax of rif:uri (iri or uri) 15:39:37 csma; next tech design topic: issue-31 15:39:55 zakim, mute me 15:39:55 MichaelKifer should now be muted 15:40:02 +q 15:40:09 zakim, unmute me. 15:40:09 Francois should no longer be muted 15:40:09 ack francois 15:40:26 csma: disjoint names for predicate symbols, function symbols and individuals 15:40:38 francois: I think we should not require them to be disjoint 15:40:51 ... 1. practical applications 15:41:15 ...2. theoretical logic: it is tradition, but no real reason for having them disjoint 15:42:17 zakim, mute me. 15:42:17 Francois should now be muted 15:42:19 q+ 15:42:25 ...3. easier for parsers and humans 15:42:38 csma: what would drawback be? 15:42:38 The antonym to disjoint-symbols is ....? overloaded symbols? 15:42:45 +q 15:43:14 chris: if they are not disjoint, many implemented systems assume they are disjoint 15:43:31 ...so will need to make a complicated encoding 15:43:40 and round-tripping will be more difficult 15:44:06 q? 15:44:29 q? 15:44:32 csma: RT more difficult because systems that required disjoint 15:44:34 ack josb 15:44:40 ...would create new symbols 15:45:20 q? 15:45:25 jos: biggest argument for having them disjoint is accessibility of the core 15:45:41 chris: it's about translation/encoding 15:45:58 s/accessibility/extensibility/ 15:46:44 chris: every rule system can express things that cannot be expressed in core 15:47:08 q? 15:47:14 ....but in this case, it would be something in the core that would be complicated to express on a lot of languages 15:47:55 csma: but if rif requires disjoint sets than a system that requires symbols that are not disjoint will have to make them disjoint for rif core 15:47:55 ack francois 15:48:01 zakim, unmute me. 15:48:01 Francois was not muted, Francois 15:48:11 When we map both s and s to s, then we cannot uniquely invert that mapping. 15:48:25 francois: want to answer to what chris said. it's a good point that RT is a problem 15:48:38 ...but we already have this problem anyway 15:49:08 ...because of all the syntactic requirements 15:49:12 +1 to Francois 15:50:29 q? 15:51:18 you cannot just rename URIs in exchange 15:51:26 zakim, mute me. 15:51:26 Francois should now be muted 15:51:33 However, sorts such as something like s and s can keep the inverse mapping to s and s unique. 15:51:33 francois: ...don't put syntactic restrictions you don't need 15:51:36 q+ 15:51:41 I agree 100% with Francois! Local variables/symbols should be allowed. 15:51:48 csma: I think francois makes a good point 15:52:02 chris: no, I think it's different 15:52:05 I think there are two conflated issues here. (1) Can you tell by the syntax of an identifier what kind of thing it is, and (2) Can you use the same string to name a predicate and a function at the same time? 15:52:15 chris: this is a semantic issue, not a syntactic one 15:52:35 +q 15:53:00 chris: (ex: rules that quantify over predicates) need different encoding, rather than just renaming things 15:53:18 s/things/symbols/ 15:53:55 q? 15:54:17 csma: question is whether we allow non-disjoint in core and dialects retrict to disjoint 15:54:36 ack harold 15:54:46 ... or whether we want core to be disjoint and dialects restrict to non-disjoint 15:54:54 francois: I put an example in the IRC 15:55:22 s/francois/harold/ 15:55:37 ack francois 15:55:45 harold: within the const element there is an atrribute "sort" where we can disambiguate 15:56:57 francois: want to answer chris' point 15:57:14 chris: I said renaming symbols is a syntactic issue 15:57:39 chris: say you have rule that quatifies over predicates, and want to translate that through RIF 15:57:56 francois: we have a misunderstanding 15:58:04 q? 15:58:37 q+ 15:59:20 francois: the position in the formula tells you what is the type of the symbol 15:59:22 Are you guys aware of something called the (typed) lambda-calculus? :-) It was invented by somebody called Alonzo Church for precisely what we are talking about. 15:59:44 ack josb 15:59:57 q+ 16:00:14 zakim, mute me. 16:00:14 Francois should now be muted 16:00:23 The language must support working with properties of properties and properties of rules. This can be done while staying first-order by having intensional semantics for predicates, where what seem to be second-order properties are really first-order properties of something in the domain of discourse which has the same name (URI) as the subject property. (This is what Pat Hayes calls "punning" and is widely implemented, although it is not normally a part of FOL 16:00:29 francois: you (chris ) are speaking of relation constants not relation variables 16:01:04 jos: a problem is that if you assert that 2 uris are equal, then they are equal only if they occur in a certain context 16:01:10 chris: is there a solution? 16:01:15 -Gary_Hallmark 16:01:25 jos: hilog provides a solution, but not acceptable for pure fol dialect 16:01:55 Higher-order logic has solved all these issues a long time ago. Are we reinventing the (square) wheel again? :-( 16:02:17 chris: if jos is right about what francois is saying, then I agree that it is a simple renaming issue 16:02:45 chris: but now what uri refers to must be interpreted contextually 16:02:46 yes 16:02:56 ack michaelk 16:02:57 zakim, unmute me 16:02:58 MichaelKifer was not muted, MichaelKifer 16:03:13 mk: i don't understand what we are discussing: I think there are 2 issues 16:03:59 1. extensibility....we are talking about pure fol dialect, so if we don't introduce sort system, then fol dialect will not be allowed 16:04:25 2. round-tripping: it will not be possible if we do not separate the symbols 16:04:37 but then the URIs are lost! 16:04:38 csma: but the separation can be done at the translation step 16:05:09 mk: I'm not sure that will really work. 16:05:47 csma: there is still an issue about round-tripping, we still haven't 16:06:01 specified exactly what we expect: will it be identical or equivalent? 16:07:05 mk: I think jos's point is that it will be be a real web language 16:07:45 jos: RT: if names of symbols (which are uris) are lost, then you 16:07:59 ...might lose some important information, and this might not be acceptable 16:08:20 mk: symbols have meaning outside of the system, so changing them might not be accepta ble 16:09:13 jos: could be a problem either way: (disjoint in rif, not in dialect) (not disjoint in rif, disjoint in dialect) 16:09:20 ...which one do we want to cater to? 16:09:54 sandro: is this the same issue as the difference between owl 1.0 and 1.1 16:10:00 chris: I think it's different 16:10:10 chris: owl doesn't translate 16:10:22 chris: here the problem occurs during translation 16:10:57 csma: I propose we keep the issue-31 open for now 16:11:14 chris: mk, did you want them disjoing or not disjoint? 16:11:23 mk: I do not have a preference 16:12:12 mk: is non dis in core, then language that do distinguish will have problems with translation from RIF 16:12:23 s/is non/if non/ 16:12:48 csma: any rule lang will have to have a way to check whether a specific rule set can be translated 16:12:53 ...and this will be one of those checks 16:13:35 chris: csma, is this a problem for production rules? 16:13:46 ...(to have non-disjoint sets) 16:13:57 csma: I don't think it has an impact on production rules 16:14:11 chris: so functions, predicates, and individuals can have the same names? 16:14:44 hassan: yes, java allows you do to this. 16:15:01 sandro: which is restricted form? 16:15:11 chris: having them disjoint is restricted form 16:15:24 +1 for Hassan's remark! 16:15:35 sandro: can something be both a property and a function? 16:15:47 as soon as you have equality, it is no longer trivial from a logical point of view 16:16:15 ...or can it only be one, and you have to disambiguate? disambiguation is too inconvenient 16:17:06 Couldn't we optionally do static analysis including automatic contextual disambiguation of s into s vs. s, changing a ruleset in this way throughout, and then translate that to systems with disjoint s and s? 16:17:12 sandro: context for equality is always individuals, but users may have a hard time keeping track of that 16:17:13 q+ 16:17:21 ack harold 16:17:40 harold: maybe we need more attention to static analysis of our rif rule sets 16:18:21 csma: disambiguation comes from sort mechanisms 16:18:50 harold: put it in translator so that user does not have to be aware of it 16:19:21 csma: can someone summarize the positions we have discussed? 16:19:46 csma: who would like to do this? 16:19:47 s/translator/validator and translator/ 16:19:51 zakim, mute me 16:19:51 MichaelKifer should now be muted 16:20:06 csma: chris, can you summarize this discussion about disjointness - in an email? 16:20:42 chris: I can start 16:20:51 sandro: how about doing it on a wiki page? 16:21:28 q+ 16:21:40 ack harold 16:22:07 harold: I think we should not edit other people's entries 16:22:32 sandro: (unless you're sure it won't offend the author) 16:22:57 csma: we will not discuss issue-25 today 16:23:09 zakim, next item 16:23:09 agendum 4. "UCR" taken up [from ChrisW] 16:23:33 csma: quick look at RIF-RAF actions 16:24:07 zakim, move to item 5 16:24:07 agendum 5. "RIFRAF" taken up [from ChrisW] 16:24:15 action-175 complete 16:24:53 csma: axel had action, but he is not here 16:25:17 csma: we still need section 5 from Leora 16:25:23 q? 16:25:29 csma: any other business? 16:25:29 Thought on 'Etiquette Rules' for Sandro's "Pros and Cons Wikis" (action on Chris): Don't start new row without need, rather align your enter your pros and cons into 'the other column' of existing rows. 16:25:31 +1 16:25:34 quit. 16:25:35 +1 16:25:42 -Hassan_Ait-Kaci 16:25:45 -Francois 16:25:46