15:54:27 <RRSAgent> RRSAgent has joined #rif
15:54:28 <RRSAgent> logging to http://www.w3.org/2008/11/14-rif-irc
15:54:34 <ChrisW> rrsagent, make logs public
15:54:41 <ChrisW> rrsagent, make minutes
15:54:41 <RRSAgent> I have made the request to generate http://www.w3.org/2008/11/14-rif-minutes.html ChrisW
15:54:45 <ChrisW> rrsagent, make logs public
15:55:00 <ChrisW> Meeting: RIF PS Task Force Telecon
15:55:09 <ChrisW> zakim, this will be rif
15:55:09 <Zakim> ok, ChrisW; I see SW_RIF()11:00AM scheduled to start in 5 minutes
15:55:27 <ChrisW> zakim, list agenda
15:55:27 <Zakim> I see 2 items remaining on the agenda:
15:55:28 <Zakim> 9. PICK SCRIBE [from ChrisW]
15:55:28 <Zakim> 10. AOB [from ChrisW]
15:55:33 <ChrisW> zakim, clear agenda
15:55:33 <Zakim> agenda cleared
15:59:48 <sandro> Zakim, who is here?
15:59:48 <Zakim> apparently SW_RIF()11:00AM has ended, sandro
15:59:50 <Zakim> On IRC I see RRSAgent, ChrisW, Harold, sandro, Zakim, trackbot
15:59:56 <sandro> zakim, this is rif
15:59:56 <Zakim> sandro, I see SW_RIF()11:00AM in the schedule but not yet started.  Perhaps you mean "this will be rif".
16:00:03 <sandro> zakim, this will be rif
16:00:03 <Zakim> ok, sandro; I see SW_RIF()11:00AM scheduled to start now
16:00:09 <Zakim> SW_RIF()11:00AM has now started
16:00:16 <Zakim> +Sandro
16:02:11 <StellaMitchell> StellaMitchell has joined #rif
16:02:34 <Zakim> +Harold
16:03:04 <LeoraMorgenstern> LeoraMorgenstern has joined #rif
16:03:08 <Zakim> +Stella_Mitchell
16:03:15 <Zakim> +Hassan_Ait-Kaci
16:03:25 <mkifer> mkifer has joined #rif
16:03:39 <Zakim> +LeoraMorgenstern
16:03:41 <sandro> ChrisW?
16:03:51 <Harold> zakim, who is on the phone?
16:03:51 <Zakim> On the phone I see Sandro, Harold, Stella_Mitchell, Hassan_Ait-Kaci, LeoraMorgenstern
16:04:04 <Hassan> Hassan has joined #rif
16:04:08 <Zakim> +Michael_Kifer
16:05:18 <sandro> ChrisW??
16:05:42 <LeoraMorgenstern> I can't hear anything; is anybody talking?
16:05:45 <Zakim> +ChrisW
16:06:00 <ChrisW> zakim, who is on the phone?
16:06:00 <Zakim> On the phone I see Sandro, Harold, Stella_Mitchell, Hassan_Ait-Kaci (muted), LeoraMorgenstern, Michael_Kifer, ChrisW
16:06:32 <ChrisW> zakim, pick a scribe
16:06:32 <Zakim> Not knowing who is chairing or who scribed recently, I propose Harold
16:07:00 <ChrisW> zakim, who is on the phone?
16:07:00 <Zakim> On the phone I see Sandro, Harold, Stella_Mitchell, Hassan_Ait-Kaci (muted), LeoraMorgenstern, Michael_Kifer, ChrisW
16:07:02 <StellaMitchell> I can
16:07:03 <ChrisW> zakim, pick a scribe
16:07:04 <Zakim> Not knowing who is chairing or who scribed recently, I propose Sandro
16:07:17 <ChrisW> Scribe: StellaMitchell
16:08:24 <StellaMitchell> Chris: There are 2 kinds of issues with PS
16:08:29 <StellaMitchell> ...1. presentability
16:08:51 <StellaMitchell> ...2 ambiguity which makes parsing difficult
16:09:18 <StellaMitchell> ... Hassan's email about difficulties he had encountered with parsing.
16:09:24 <StellaMitchell> ...5 points
16:10:08 <StellaMitchell> Hassan:  issue 1 (from email).  This has been discussed in previous emails also.
16:10:13 <sandro> q+
16:10:59 <StellaMitchell> ...Hassan proposes to put IRI's in quotes in Prefix and Base directives
16:11:18 <StellaMitchell> Sandro:  you could use white spaces as a delimiter
16:11:44 <ChrisW> Prefix         ::= 'Prefix' '(' Name IRI ')'
16:12:10 <StellaMitchell> Hassan:  it's better to make a context-free tokenizer
16:13:11 <Hassan> Prefix         ::= 'Prefix' '(' Name STRING ')'
16:13:51 <sandro> <http:...>
16:13:57 <StellaMitchell> mk: why can't we use an IRI constant here?  with anglebrackets
16:14:18 <StellaMitchell> hassan: ok, can use angle brackets instead of quotes as delimiters
16:15:47 <StellaMitchell> sandro: we should either use quotes or anglebrackets, and will decide which later
16:16:17 <StellaMitchell> Chris: issue 2.  minus sign in identifiers
16:17:09 <StellaMitchell> ->  :-
16:17:27 <sandro> sandro: lets disallow hyphen in identifiers
16:17:47 <sandro> switch to: camelcase or underscore
16:18:01 <StellaMitchell> hassan,  you don't have a problem with the underscore for localnames?
16:19:10 <Hassan>  In the RIF specification of the EBNF the Rule Language, it is specified that:
16:19:11 <Hassan>   IRIMETA        ::= '(' IRICONST? (Frame | 'And' '(' Frame ')')? ')'
16:19:11 <Hassan>   Frame          ::= TERM '[' (TERM '->' TERM) ']'
16:19:11 <Hassan>   TERM           ::= IRIMETA? (Const | Var | Expr | 'External' '(' Expr ')')
16:19:11 <Hassan>   Const          ::= '"' UNICODESTRING '"^^' SYMSPACE | CONSTSHORT
16:19:11 <Hassan>   SYMSPACE       ::= ANGLEBRACKIRI | CURIE
16:19:14 <Hassan>   
16:19:16 <Hassan> where CONSTSHORT, ANGLEBRACKIRI, and CURIE are defined (in the DTB shorthand notation for RIF constants) by:
16:19:17 <Hassan>   CURIE         ::= PNAME_LN | PNAME_NS
16:19:20 <Hassan>   CONSTSHORT    ::= ANGLEBRACKIRI         // shorthand for "..."^^rif:iri
16:19:21 <Hassan>                   | CURIE                 // shorthand for "..."^^rif:iri
16:19:23 <Hassan>                   | '"' UNICODESTRING '"' // shorthand for "..."^^xs:string
16:19:25 <Hassan>                   | NumericLiteral        // shorthand for "..."^^xs:integer,xs:decimal,xs:double
16:19:27 <Hassan>                   | '_' LocalName         // shorthand for "..."^^rif:local
16:19:30 <Hassan>   
16:19:31 <Hassan> where:
16:19:32 <sandro> I suggest using http://www.w3.org/TeamSubmission/turtle/ whereever reasonable.
16:19:33 <Hassan>   ANGLEBRACKIRI ::= '<' ([^<>"{}|^`\]-[#x00-#x20]) '>'
16:19:35 <Hassan>   PNAME_LN      ::= PNAME_NS PN_LOCAL
16:19:37 <Hassan>   PNAME_NS      ::= PN_PREFIX? ':'
16:19:39 <Hassan>   PN_LOCAL      ::= (PN_CHARS_U | [0-9]) ((PN_CHARS|'.') PN_CHARS)?
16:19:41 <Hassan>   PN_PREFIX     ::= PN_CHARS_BASE ((PN_CHARS|'.') PN_CHARS)?
16:19:43 <Hassan>   PN_CHARS_U    ::= PN_CHARS_BASE | '_'
16:19:45 <Hassan>   PN_CHARS      ::= PN_CHARS_U
16:19:47 <Hassan>                   | '-'
16:19:50 <Hassan>                   | [0-9]
16:19:51 <Hassan>                   | #x00B7
16:19:54 <Hassan>                   | [#x0300-#x036F]
16:19:55 <Hassan>                   | [#x203F-#x2040]
16:19:58 <Hassan>   PN_CHARS_BASE ::= [A-Z]
16:19:59 <Hassan>                   | [a-z]
16:20:01 <Hassan>                   | [#x00C0-#x00D6]
16:20:03 <Hassan>                   | [#x00D8-#x00F6]
16:20:05 <Hassan>                   | [#x00F8-#x02FF]
16:20:08 <Hassan>                   | [#x0370-#x037D]
16:20:09 <Hassan>                   | [#x037F-#x1FFF]
16:20:11 <Hassan>                   | [#x200C-#x200D]
16:20:13 <Hassan>                   | [#x2070-#x218F]
16:20:16 <Hassan>                   | [#x2C00-#x2FEF]
16:20:17 <Hassan>                   | [#x3001-#xD7FF]
16:20:20 <Hassan>                   | [#xF900-#xFDCF]
16:20:21 <Hassan>                   | [#xFDF0-#xFFFD]
16:20:23 <Hassan>                   | [#x10000-#xEFFFF]
16:20:25 <Hassan>   
16:21:07 <StellaMitchell> hassan: DTB points to other specs....parts of the above grammar are from those referenced specs
16:22:05 <StellaMitchell> Chris: the - is allowed in the curie notation?
16:23:13 <StellaMitchell> Chris: we are inheiriting dash inside identifiers from curie syntax
16:23:30 <StellaMitchell> we have dash in our built-in names also
16:24:05 <StellaMitchell> sandro: we will inheirit additional difficult things from curie grammar also
16:24:50 <StellaMitchell> hassan:  I'm not aiming my prototype to cover all posssible valid cases....but rather most of them
16:25:10 <StellaMitchell> sandro: we could give a restricted definition of identifiers
16:26:19 <sandro> http://www.w3.org/TeamSubmission/turtle/#name
16:26:33 <StellaMitchell> chris: then we have to write our own syntax for these, and not use curie
16:26:50 <StellaMitchell> sandro: could possible use the one above
16:27:11 <sandro> nameStartChar | '-' | [0-9] | #x00B7 | [#x0300-#x036F] | [#x203F-#x2040]
16:27:16 <sandro> (turtle)
16:27:31 <StellaMitchell> axel points to sparql grammar
16:27:51 <StellaMitchell> Chris: what if we remove dash from other places, outside of identifiers
16:28:35 <StellaMitchell> sandro: I can't image readable rules without infix subtraction
16:28:44 <StellaMitchell> s/image/imagine/
16:28:55 <ChrisW> ex:-
16:28:57 <ChrisW> -
16:29:38 <Hassan> foo-bar
16:29:48 <sandro> External(func:numeric-subtract(2 1)) 	2 - 1 	"+" "-" "*" "/" as in programming languages
16:30:02 <StellaMitchell> you could expand all "shortcuts" before parsing
16:30:05 <sandro> (quoting http://www.w3.org/TR/rif-ucr/)
16:30:13 <Harold> The "-" in the abridged 17-4 would not be part of an identifier.
16:31:39 <StellaMitchell> but without the infix operator which is not currently part of the grammar, you can preprocess
16:31:43 <Harold> So omitting "-" from identifiers helps introducing abridged PS.
16:32:54 <StellaMitchell> sandro: so can we remove - from identifers
16:33:03 <StellaMitchell> chris: then we have to reproduce syntax for curies
16:33:32 <StellaMitchell> hassan: if minus signs are allowed in identifiers in turtle, etc and they allow infix operators, how do they handle that?
16:33:46 <StellaMitchell> chris: they don't have - as a separate operator
16:34:39 <StellaMitchell> chris: I think it's more important to be in line with the other standards, than it is to have pretty substraction syntax
16:35:09 <Harold> External(func:subtract-dateTimes(?deliverydate ?scheduledate))   OR  External(?deliverydate-?scheduledate)
16:35:44 <StellaMitchell> Chris: let's think about this one, and come back to it
16:36:39 <Harold> External(func:subtract-dateTimes(?deliverydate ?scheduledate))   OR  External(?deliverydate - ?scheduledate)
16:36:46 <StellaMitchell> sandro: another option (that others do) is to require spaces around the - when used as substraction
16:37:01 <StellaMitchell> then pre-processing could handle the space separated one too
16:37:23 <StellaMitchell> Chris: issue 4
16:37:51 <StellaMitchell> ...arguments of uniterms are either all attributed or all positional
16:38:46 <StellaMitchell> hassan: for bottom up parser, a problem
16:38:56 <StellaMitchell> is this because of shortcuts again?
16:39:49 <StellaMitchell> or, could you also pre-process away curie notation?
16:41:36 <StellaMitchell> hassan: could allow mixing at the ebnf level, and then check for additional syntax errors (mixing) later?
16:42:02 <StellaMitchell> is it just curies that cause this ambiguity?
16:42:39 <StellaMitchell> a name does already does look different than a const
16:43:19 <StellaMitchell> chris: confusing that syntax for named arguments in uniterms is the same as syntax for slots in frames
16:43:40 <sandro>    foo( [color]red )     would work
16:45:22 <StellaMitchell> color is clearly not a const anyway, since it isn't in double quotes
16:45:32 <ChrisW> foo[ color::red size::big ]
16:45:36 <sandro>    foo( color : red )     would work
16:45:49 <sandro> (solving Chris's problem, not Hassan.)
16:46:10 <StellaMitchell> sandro: we are discussing 2 differernt issues about named args
16:46:22 <StellaMitchell> ...hassan's re parsing, and chris's re readability
16:46:47 <StellaMitchell> why do you need extra delimiters here?
16:46:51 <ChrisW> foo{ color::red size::big }
16:47:07 <StellaMitchell> consts (cannonical) already are delimited by quites
16:47:14 <StellaMitchell> quotes
16:47:19 <ChrisW> foo! color::red size::big !
16:48:22 <sandro> foo((a->b))
16:48:30 <sandro> f((a)->b)
16:48:43 <sandro> foo( named color->red)
16:49:11 <sandro> foo(NAMED color->red)
16:49:16 <StellaMitchell> sandro, why do you think color can be confused as posiitinal argument??
16:49:47 <sandro> StellaMitchell, it's not truly ambiguous, it just requires more look-ahead.
16:49:57 <StellaMitchell> why does it require more lookahead?
16:49:58 <mkifer> foo(! color-> red !)
16:50:06 <StellaMitchell> consts have to be in quotes
16:50:22 <StellaMitchell> and that NAMED will make it context sensitive parsing
16:50:50 <Harold> We implemented this in ANTLR.
16:51:34 <StellaMitchell> I don't see why you need arbitrary lookahead
16:52:46 <sandro>  foo(-> color -> red ->)      :-)  :-)
16:54:23 <StellaMitchell> cw: useful for readability to distinguish named arg predicates from frames
16:55:16 <StellaMitchell> I don't understand why it requires more lookahead
16:55:43 <StellaMitchell> i.e. why parsing can't be simple with expanded consts
16:55:57 <StellaMitchell> I just typed it
16:56:35 <StellaMitchell> that is only in the case of curies?
16:56:46 <StellaMitchell> so if you expanded them, it would go away?
16:57:42 <Hassan> foo ( bar baz -> foo boo -> goo )
16:58:11 <StellaMitchell> that is not correct in many ways
16:58:16 <StellaMitchell> you would fail at the first foo
16:58:57 <sandro> foo ( color -> red name -> rover )     ?  
17:00:12 <Harold> PS is Human-Oriented Syntax.
17:00:28 <Harold> So it's ok if parsers have some harder time.
17:00:36 <StellaMitchell> why not add commas?
17:01:01 <Harold> After all, parsing is done only once, before the machine has to start its real (deductive) work..
17:01:48 <Harold> It's not on the XML level.
17:02:20 <StellaMitchell> I don't know why identifier would be in the lookaead set in 82
17:02:39 <StellaMitchell> Chris: issue 5
17:03:51 <StellaMitchell> Hassan: ....cannot tell the difference between special case and general case
17:03:54 <LeoraMorgenstern> I will leave as soon as my 12 noon call starts (it's running late), but I am just listening anyway.
17:05:01 <StellaMitchell> ...the ambiguity is that you have a grammar that gives general rules and then within that grammar there is another grammar that gives more specific rules
17:05:16 <StellaMitchell> mk: where is the difficulty? is it the ^^?
17:05:38 <StellaMitchell> hassan:  we are talking about the rif:iri suffix
17:06:37 <StellaMitchell> ...already past the "..."^^curie
17:06:52 <StellaMitchell> ... I am talking about specifically rif:iri that complicates things at the lexical level
17:07:35 <StellaMitchell> sandro:  I think a RIF parser should never be looking inside an IRI?
17:07:45 <StellaMitchell> Chris: an iri or a string
17:07:52 <StellaMitchell> s/string/string?/
17:07:58 <StellaMitchell> sandro: either
17:08:54 <StellaMitchell> Hassan:  at the lexical level we require that it be rif:iri when it is an iri const complicates lexical parsing
17:09:33 <StellaMitchell> ...it would be easier to accept any curie there and check that it's rif:iri later
17:10:08 <StellaMitchell> I will send an email example to make the problem more clear
17:10:55 <StellaMitchell> Chris: another meeting same time next week?
17:11:09 <StellaMitchell> ... works for everyone
17:11:40 <StellaMitchell> mk: we haven't discussed the escape symbols...i.e. for double quotes within double quoted strings
17:11:45 <Zakim> -Sandro
17:11:47 <Zakim> -LeoraMorgenstern
17:11:47 <Zakim> -Michael_Kifer
17:11:48 <Zakim> -Hassan_Ait-Kaci
17:11:48 <Zakim> -ChrisW
17:11:50 <Zakim> -Harold
17:12:00 <Zakim> -Stella_Mitchell
17:12:02 <Zakim> SW_RIF()11:00AM has ended
17:12:03 <Zakim> Attendees were Sandro, Harold, Stella_Mitchell, Hassan_Ait-Kaci, LeoraMorgenstern, Michael_Kifer, ChrisW
17:12:07 <ChrisW> rrsagent, make minutes
17:12:07 <RRSAgent> I have made the request to generate http://www.w3.org/2008/11/14-rif-minutes.html ChrisW
18:29:28 <Zakim> Zakim has left #rif