IRC log of rdfcore on 2002-11-22

Timestamps are in UTC.

14:45:28 [RRSAgent]
RRSAgent has joined #rdfcore
14:45:33 [DanConn]
Zakim, who is on the phone?
14:45:34 [Zakim]
sorry, DanConn, I don't know what conference this is
14:45:35 [Zakim]
On IRC I see RRSAgent, DanConn, Zakim, AaronSw, logger
14:45:40 [DanConn]
Zakim, this will be RDF
14:45:41 [Zakim]
ok, DanConn
14:49:01 [bwm]
bwm has joined #rdfcore
14:49:16 [bwm]
morning all
14:49:28 [AaronSw]
14:53:03 [jang_scri]
jang_scri has joined #rdfcore
14:53:27 [AaronSw]
already wearing the scribe hat. look at that enthusiasm!
14:53:39 [jang_scri]
look at last week's minutes :-/
14:53:48 [bwm]
G'day Jan - are you really scribing again?
14:53:51 [DaveB]
DaveB has joined #rdfcore
14:54:01 [jang_scri]
happy to, dialing in, however...
14:54:18 [jang_scri]
on a laptop with no mouse, so if someone (aaron) can paste urls into the transcipt that'll make my life easier
14:54:19 [Zakim]
SW_RDFCore()10:00AM matches both RDF and rdf
14:54:38 [AaronSw]
zakim, this is rdf
14:54:39 [Zakim]
ok, AaronSw
14:54:44 [Zakim]
14:54:45 [Zakim]
SW_RDFCore()10:00AM has ended
14:54:48 [DanConn]
er... jan, did you put yourself down as scribe on purpose? I don't think a mistake in the minutes one week should count as volunteering for the next week.
14:55:13 [Zakim]
SW_RDFCore()10:00AM matches both RDF and rdf
14:55:31 [bwm]
zakim, who is on the phone?
14:55:31 [AaronSw]
zakim, who is on the phone?
14:55:32 [Zakim]
I notice SW_RDFCore()10:00AM has restarted
14:55:32 [Zakim]
On the phone I see ??P3
14:55:33 [Zakim]
On the phone I see ??P3
14:55:39 [Zakim]
14:55:42 [Zakim]
14:55:43 [Zakim]
14:55:51 [jang_scri]
zakim, ??p4 is ilrt
14:55:52 [Zakim]
+Ilrt; got it
14:56:00 [jang_scri]
zakim, ilrt has jang, daveb
14:56:02 [Zakim]
+Jang, Daveb; got it
14:56:04 [bwm]
zakim ??P3 is bwm
14:56:54 [DanCon]
DanCon has joined #rdfcore
14:57:02 [Zakim]
14:57:08 [DaveB]
14:57:29 [bwm]
that's a rather drastic way to mute
14:57:32 [Zakim]
14:57:36 [Zakim]
14:57:37 [Zakim]
14:57:45 [DaveB]
Zakim, ??p4 is ilrt
14:57:46 [Zakim]
+Ilrt; got it
14:58:48 [bwm]
zakim, who is on the phone?
14:58:49 [Zakim]
On the phone I see ??P3, Ilrt
14:59:12 [bwm]
Zakim, ??P3 is bwm
14:59:13 [Zakim]
+Bwm; got it
14:59:20 [jjc]
jjc has joined #rdfcore
14:59:24 [Zakim]
14:59:26 [DanCon]
FYI: EricM has no network this morning, I gather
14:59:39 [Zakim]
14:59:40 [jjc]
what's the meeting code?
14:59:44 [jang_scri]
14:59:45 [bwm]
14:59:58 [jang_scri]
gotta get up in the morning to beat me, bwm
15:00:06 [jang_scri]
early! I mean, ea
15:00:06 [bwm]
zakim, who is on the phone?
15:00:07 [Zakim]
On the phone I see Bwm, Ilrt, PatH, AaronSw
15:00:08 [jang_scri]
rly in the morning
15:00:35 [Zakim]
15:00:50 [AaronSw]
zakim, ??P7 is stevep
15:00:51 [Zakim]
+Stevep; got it
15:00:51 [Zakim]
15:01:02 [AaronSw]
zakim, ??P8 is jjc
15:01:03 [Zakim]
+Jjc; got it
15:01:21 [Zakim]
15:01:28 [Zakim]
15:01:50 [Zakim]
15:02:12 [AaronSw]
zakim, ??p10 is patricks
15:02:13 [Zakim]
+Patricks; got it
15:02:44 [jang_scri]
we were hoping you'd be halfway up a mountain or something exciting like that, patrick!
15:03:05 [Zakim]
15:03:24 [Zakim]
15:03:34 [AaronSw]
zakim, ??p12 is josd
15:03:35 [Zakim]
+Josd; got it
15:03:52 [jang_scri]
jang to scribe this week
15:03:54 [jang_scri]
next week:
15:04:11 [JosD]
JosD has joined #rdfcore
15:04:17 [jang_scri]
next meeting... thanksgiving next week
15:04:28 [Zakim]
+ +1.617.253.aaaa
15:04:29 [jang_scri]
sounds like quorate next week
15:04:32 [jang_scri]
next telecon, next week
15:04:34 [jang_scri]
jjc to scribe
15:04:39 [AaronSw]
zakim, +1.61 is danbri
15:04:40 [Zakim]
+Danbri; got it
15:04:55 [jang_scri]
roll call
15:04:57 [bwm]
zakim, who is on the phone
15:04:59 [Zakim]
I don't understand 'who is on the phone', bwm
15:05:05 [bwm]
zakim, who is on the phone?
15:05:06 [Zakim]
On the phone I see Bwm, Ilrt, PatH, AaronSw, Stevep, Jjc, EricM, Patricks, FrankM, DanC, Josd, Danbri
15:05:15 [danbri]
danbri has joined #rdfcore
15:05:22 [jang_scri]
zakim, ilrt has daveb jang
15:05:23 [Zakim]
+Daveb, Jang; got it
15:05:42 [jang_scri]
regrets: mike dean.
15:05:54 [jang_scri]
15:06:19 [jang_scri]
(jjc's range, domain move to under test cases)
15:06:30 [jang_scri]
minutes of the last meeting:
15:06:34 [jang_scri]
15:06:39 [jang_scri]
completed actions:
15:06:40 [jang_scri]
all done
15:07:01 [DaveB]
jang test cases:
15:07:05 [DaveB]
working on the actions...
15:07:09 [DaveB]
... main thing is DT test cases
15:07:17 [DaveB]
... working on these
15:07:27 [DaveB]
see url in agende item 7
15:08:04 [DaveB]
jang_scri: issues are
15:08:04 [AaronSw]
15:08:15 [DaveB]
... need to get an answer to whether lang is in or not
15:08:42 [DaveB]
bwm: what did we decide re lang component in syntactic component in literal
15:08:51 [DaveB]
DanCon: what did record show?
15:08:51 [DanCon]
which test case, jan?
15:09:15 [DaveB]
jang_scri: when not doing DT-entail, understood we had tidy nodes with
15:09:29 [DaveB]
... lang and string
15:09:46 [bwm]
15:10:16 [AaronSw] doesNotEntail
15:10:40 [bwm]
15:10:43 [jang_scri]
ACTION me to slap frag ids into that
15:11:12 [DaveB]
one is a premise, one a conclusion. test is a non entailment
15:11:37 [DanCon]
does { <">> <">> "10"^^<">> .} entail { <> <> "10"@fr^^<> .} ?
15:11:58 [DaveB]
normal rdf entailment here
15:12:02 [DaveB]
so saying lang is important here
15:12:23 [DanCon]
which part of the model theory spec is relevant, please?
15:12:36 [DaveB]
restating test004a does not rdf-entail test004b
15:12:52 [DaveB]
(will discuss dt-entail in a bit)
15:12:54 [DanCon]
15:13:07 [DaveB]
path quotes above
15:13:19 [DaveB]
plain literal is the pair (string, lang)
15:13:39 [DaveB]
jang_scri: 2nd half is ...
15:13:50 [DaveB]
... when you do dt-entail and have knowledge of the DT, in this case xsd integer
15:13:57 [DanCon]
which test are we talking about now?
15:14:01 [DaveB]
... then you have the entailemnt of teach other
15:14:16 [DaveB]
so, saying test004a dt-entails test004b
15:14:25 [DaveB]
15:14:29 [DaveB]
^ the test id
15:14:50 [AaronSw]
15:14:54 [JosD]
15:15:08 [DaveB]
^ above is the (404 at present) uri of the test case in discussion
15:15:12 [DaveB]
... with the dt-entailment
15:15:24 [AaronSw]
<test:datatypeSupport rdf:resource="" />
15:15:29 [DaveB]
jang_scri: the manifest lists the entailment rules
15:15:31 [AaronSw]
<test:entailmentRules rdf:resource="" />
15:15:32 [AaronSw]
15:16:16 [DaveB]
jjc: suggests using the uri of the test in the manifest
15:16:30 [DaveB]
ACTION jang: change the test uri to point to datatypes/Manifest#test-id
15:17:27 [DaveB]
DanCon: do we need schema rules here?
15:17:33 [DaveB]
jang_scri: no, could delete that
15:17:58 [DaveB]
ACTION jang: delete schema rule from datatypes/Manifest.rdf
15:18:19 [DaveB]
PatH: defer language to the datatype authority
15:18:25 [DaveB]
such as in XSD
15:18:43 [DaveB]
jjc: concepts makes it clear that language is not important except for rdf:XMLLiteral
15:18:47 [bwm]
15:18:53 [DaveB]
... in Datatyping
15:19:10 [DaveB]
jjc: maybe need to tie MT to the above
15:19:42 [DaveB]
ACTION PatH: Tie MT datatype to the sec that points out language isn't important to DT entailment except for rdf:XMLLiteral
15:19:54 [DaveB]
above ACTION also to jjc
15:21:48 [DaveB]
discussion of XSD use string to value mappings, language not used
15:24:10 [AaronSw]
hm, xsd has a language tag datatype
15:24:50 [DaveB]
jjc: I couldn't work out what action you took
15:25:06 [jjc]
ACTION: jjc create daattype test case showing language is not relevant.
15:25:38 [jjc]
ACTION: jjc check RDF Concepts does not allow a synonym for rdf:XMLLiteral
15:26:13 [jang_scri]
question arises about XMLLiteral "sutypes"
15:26:15 [jang_scri]
subtypes even
15:26:48 [jang_scri]
more recent drafts have a more coherent story on "what is a literal"
15:27:05 [jang_scri]
we have one special case, but that's better than the three disctinct cases we had before
15:27:38 [DanCon]
yes, another test case for subtypes of XMLLiteral, please.
15:27:40 [jang_scri]
pats: so we need some words somewhere on this
15:28:14 [jang_scri]
path: subclassof doesn't establish a relationship between DATATYPES
15:28:37 [jang_scri]
jjc: there's j borden's worries about wanting different classes of xml lits
15:29:14 [jang_scri]
path: we could say, in all cases the lang tag is passed to the dt, the dt can ignore it or not
15:30:22 [jang_scri]
jang_scri: not knowing about an dt (don't put it in test case) you can catch this with an entilament
15:30:30 [jang_scri]
bwm: what about i18n guys?
15:30:42 [jang_scri]
jjc: we came to a compromise: it's in there, doesn't mean much.
15:31:06 [jang_scri]
jjc: we have a warning against confusing language with locale
15:31:14 [jang_scri]
path: a literal is a string + (optional) langtag
15:31:19 [jang_scri]
a dt lit is a simple lit + dt
15:31:41 [jang_scri]
the DT gets the whole simple literal; it can ignore part of the pair (as, we note, xsd always does)
15:32:13 [jang_scri]
pats: we can note also that xmlliteral subtypes take the lang tag
15:32:24 [jang_scri]
JosD: this is a nightmare
15:32:42 [jang_scri]
it's level-mixing, not clean, it's ugly and awkward
15:32:54 [jang_scri]
path: won't this tidy that up?
15:33:02 [jang_scri]
JosD: no, I've been folloowing that route this week.
15:33:22 [jang_scri]
if you rely on the dt and you don't know it, how can you make an inference engine make progress?
15:33:54 [jang_scri]
jjc: can we _discourage_ the use of lang tags?
15:34:09 [jang_scri]
JosD: why not define language-driven datatypes?
15:34:28 [jang_scri]
putting it in the realm of the literal isn't a good thing in my opinion
15:34:42 [jang_scri]
JosD: then I can live with status quo
15:35:03 [DanCon]
could somebody write down the proposal bwm's talking about?
15:35:09 [jang_scri]
JosD: simplification is always good
15:35:45 [jang_scri]
pats: another option: simply say that languge is a scoping issue, has no real business in any literals
15:36:03 [DanCon]
what pats is saying is tantamount to a request to re-open rdfms-literal
15:36:06 [jang_scri]
and we fix m&s by saying that xml:lang is relevant to rdf parser, but not to an application looking at the graph
15:36:11 [jang_scri]
DaveB: no
15:36:34 [jang_scri]
jjc: also disagree; there was a decision about this ages ago.
15:36:47 [jang_scri]
bwm: restates proposal
15:36:59 [jang_scri]
to relax the constrain that dts cannot take the lang tag into account in the mapping
15:37:29 [DanCon]
no, it's NOT A PAIR. strings are just strings.
15:38:15 [Zakim]
15:38:17 [jang_scri]
(round table on who likes this)
15:38:32 [Zakim]
15:38:53 [jang_scri]
pats: I still think having a special case with xmlliteral is a mistake...
15:39:05 [jang_scri]
ie, xmlliterals are legacy, not dt'ed literals
15:39:31 [jang_scri]
DanCon: I've sympathy for that, but we've decided on our current state: what's the new information
15:39:39 [jang_scri]
pats: conflict with subclass semantics
15:39:45 [jang_scri]
bwm: not if we do it right
15:39:58 [DanCon]
FIRST: which decision(s) do you want to reopen. THEN: what's the new information?
15:40:36 [jang_scri]
bwm: we have the entailment test as it is now, you need to know what the dts are.
15:40:49 [jang_scri]
so we DON't want an dt test case
15:41:37 [jang_scri]
bwm: case ofa subclass of xmlliteral: surely a subtype of xmlliteral can take into account the
15:41:43 [jang_scri]
15:41:55 [jang_scri]
DanCon: let jjc write the test case, we vote next week.
15:42:13 [jang_scri]
(jjc describes proposed test case to jos)
15:43:25 [jang_scri]
path: an rdf engine can post an error if it finds a DT it doesn't grok
15:43:32 [jang_scri]
path: I used the term "reasoner"
15:44:20 [jang_scri]
ACTION danc: review the semantics document, with a view to language, not software specification
15:44:21 [DanCon]
"reasoner" does not occur in
15:44:52 [jang_scri]
ACTION jjc to generate the dt test case with dt
15:45:03 [DanCon]
ah; "RDF reasoner" does occur in
15:46:06 [DaveB]
jang_scri: other iss ue swhere you know about datatypes and their wellformedness...
15:46:07 [DanCon]
(I'm seriously considering re-opening rdfs:format, jjc; I have informed the chair)
15:46:26 [DaveB]
... whether an rdf document that contains a badlyf ormatted liter should create an rdf graph, is an error or what?
15:46:57 [DaveB]
PatH: if it is an illformed literal, rdf graph is legal but denotes a value that you know is not a literal
15:47:17 [DaveB]
jang_scri: will go back to think about this - make it a non entailment
15:47:52 [DaveB]
PatH: anotehr case is the DT clash case when it is well formedl, known to have a value but you have a range ... [lost it]
15:48:12 [DaveB]
ACTION jang: take a copy of sketch of the abvoe case re DT calashing in the MT and make a test case
15:48:34 [DaveB]
jang_scri: an XSD test - semantic-queiv-bnetween-datatypes
15:48:48 [bwm]
15:48:55 [DaveB]
test005a.nt implies test005b.nt
15:49:11 [DaveB]
jang_scri: propose to not approve it sincfe more to do with semantics of XSD
15:49:51 [danbri]
15:49:57 [jang_scri]
jjc: likes it
15:50:24 [jang_scri]
pats: what about creating a subtypes of one DT to another and using that
15:50:25 [DanCon]
??? "datatype X is a subclass of datatype Y" ??? what does PatS mean by that?
15:50:27 [bwm]
ack danbri
15:50:54 [danbri]
15:51:00 [bwm]
ack dancon
15:51:02 [Zakim]
DanCon, you wanted to ask about an action to ask the I18N WG about our resolution on lang, literals, and datatypes
15:52:13 [jang_scri]
ACTION jjc: check with i18n guys on whether they';re happy with the datatypes as they are
15:52:26 [jang_scri]
bwm: I like patricks example
15:52:42 [jang_scri]
DanCon: if soneone asked about this on the comments list, we owe them an answer
15:53:31 [jang_scri]
jjc: don't argue about mail messages, argue about the spec!
15:53:49 [jang_scri]
DanCon: xsd:integer has a value space that's the integers.
15:54:36 [jang_scri]
pats: the _answer_ is yes or no, correct, but we have to say what it means to say "datatype support"
15:54:36 [jjc]
15:54:38 [Zakim]
15:54:55 [jang_scri]
pats: it says you have support for xsd:integer
15:55:01 [jang_scri]
and for xsd:decimal
15:55:09 [jang_scri]
but NOT for the mapping - if any - between them
15:55:27 [jang_scri]
DanCon: that's all you need?!
15:55:52 [jang_scri]
xsd schema spec doesn't refer to lexical->value mapping
15:56:02 [jang_scri]
15:56:06 [DaveB]
where did "If you know..." come from in that explanation. You don't know since XSD datatyping is not in RDF.
15:56:11 [jang_scri]
path: I'm still not persuaded about this
15:56:12 [danbri] [[
15:56:31 [danbri]
XML Schema 1.1
15:56:31 [danbri]
You can help
15:56:31 [danbri]
The XML Schema WG is currently working to develop a set of requirements for XML Schema 1.1, which is intended to be mostly compatible with XML Schema 1.0 and to have approximately the same scope, but also to fix bugs and make whatever improvements we can, consistent with the constraints on scope and compatibility.
15:56:33 [danbri]
15:56:34 [jang_scri]
[scribe can't keep up with this]
15:56:39 [danbri]
15:57:13 [jang_scri]
jjc: finds sections on xsd spec that deals with this.
15:57:30 [DanCon]
"The number of literals for each value has been kept small; for many datatypes there is a one-to-one mapping between literals and values." --
15:58:14 [jang_scri]
bwm: jjc is saying that this test case, int is derived from decimal
15:58:39 [jang_scri]
pats: when we say "datatype support", does that mean everything that that might mean..?
15:58:41 [jang_scri]
15:59:26 [jang_scri]
DanCon: it's not essential to know that there's some relationship between the dt mappings
15:59:36 [jang_scri]
it's only enough to know that int 10 = decimal 10.0
15:59:36 [danbri]
more from xml schema page: [[
15:59:39 [danbri]
... If you have a schema you would like to see listed here, or if you have a set of test cases for use in software development, please send an announcement to xmlschema-dev. (If you are impatient, you may also wish to send mail to Michael Sperberg-McQueen calling his attention to your email on xmlschema-dev and suggesting your materials be added to this list.)
15:59:41 [danbri]
15:59:49 [danbri] they're explicitly soliciting test cases.
16:00:36 [jang_scri]
[a scrap ensues]
16:01:12 [bwm]
ack jjc
16:02:02 [jjc]
16:02:06 [bwm]
ack jan
16:02:25 [bwm]
ack danbri
16:04:14 [Zakim]
16:04:17 [jang_scri]
jjc breaking up, redialling
16:05:19 [em]
em has joined #rdfcore
16:05:47 [jang_scri]
jang's example: datatype 1 = natural odd numbers, datatype 2 = primes; no subtype relationship, an entailment would still hold
16:06:41 [Zakim]
16:06:48 [jang_scri]
zakim, ??p6 is jjc
16:06:49 [Zakim]
+Jjc; got it
16:08:16 [jjc]
If you have xsd:integer then you have the fact that it is derived from xsd:deciaml.
16:09:54 [bwm]
ack jjc
16:10:18 [jang_scri]
xsd decimal 10.0 specifies the same thing as integer 10
16:10:35 [jang_scri]
pats: what is the definition of "data type support"
16:11:24 [jang_scri]
16:12:04 [jang_scri]
frankm: is the distinction between datatypes and cross-datatype knowhow artificial?
16:12:13 [danbri]
em, see
16:12:31 [jang_scri]
jjc: you can't implement xsd:integer without implementing xsd:decimal, since it's part of that definition
16:12:56 [DaveB]
did we capture what was "data type support"?
16:13:02 [jang_scri]
pats: relationshipts to other dts are part of the definition of that dt
16:13:22 [jang_scri]
so scheme:integer 10 = xsd:integer 10 is NOT an entailment that we would want to sanction
16:13:28 [jang_scri]
although some application may want to provide that
16:13:47 [DanCon]
eek! please, no, let's not depend on comments in test cases!
16:14:06 [DaveB]
I assume that really meant prose in the test casses WD
16:14:17 [DaveB]
rather than # comments in a single test case
16:14:20 [jang_scri]
jang meant prose in test case wd about this
16:14:23 [Zakim]
16:14:51 [jang_scri]
substitute 10.0 for 1E1 in this test case...
16:15:19 [jang_scri]
pats: happy as long as there's some text about relationshipts between datatypes
16:15:33 [jjc]
Add: "Note: xsd:integer is derived from xsd:decimal."
16:15:51 [jang_scri]
jjc: add to the desc of the test case, the words above
16:16:22 [jang_scri]
jjc: the reason why it's relevant:
16:16:33 [jang_scri]
you can't implement xsd:integer without implementing xsd:decimal
16:16:38 [DanCon]
what's not quite clear to me is what this stuff means: [[
16:16:40 [DanCon]
<test:datatypeSupport rdf:resource="" />
16:16:40 [DanCon]
<test:datatypeSupport rdf:resource="" />
16:16:40 [DanCon]
16:17:15 [jang_scri]
16:19:14 [jang_scri]
it means: you know about xsd:integer, you know about xsd:decimal, and for any subset of {xsd:integer, xsd:decimal}, you know about the relationships that hold over that subset
16:19:42 [jang_scri]
jjc: double 10 =? integer 10
16:20:00 [jang_scri]
let's accept this simpler question, then look at the harder one - where we might well disagree.
16:20:53 [DanCon]
there are three relevant tests: (a) integer/decimal, where the subtype relation is clear and explicit. (b) hex/binary, which are specified to share a value space but not to have any subtype relationship, and (c) float/decimal, which are not specified to have the same value space.
16:21:09 [jang_scri]
yes, that sounds perfect dan
16:21:17 [jang_scri]
but I don't think we'll get agreement :-/
16:21:52 [jang_scri]
pats: how can I approve this test case without the criteria that there be a defined relationship between xsd:integer and xsd:decimal
16:22:23 [jang_scri]
16:22:29 [DanCon]
please put the question, mr. chair
16:22:36 [jang_scri]
it seems to me that the consequence of that pats is saying ...
16:22:46 [jang_scri]
this issue is independent of specific test cases
16:22:57 [jang_scri]
we have to clarify what is meant by datatype entailment
16:23:10 [jang_scri]
then we'll answer every test case
16:25:46 [jang_scri]
path solves all our problems again :-)
16:25:53 [jang_scri]
this one is NOT APPROVED yet
16:26:29 [jang_scri]
16:26:29 [DanCon]
which test case?
16:26:47 [danbri]
16:26:58 [jang_scri]
semantic equivalence within a type
16:27:17 [jang_scri]
16:27:20 [danbri]
16:27:30 [jang_scri]
16:27:44 [jang_scri]
one entails the other & vice versa
16:28:27 [DaveB]
16:28:47 [jang_scri]
jang_scri: wanted to make this explicit
16:28:52 [jang_scri]
that these types are equivalent
16:29:19 [jang_scri]
although I sympathise that this might cause problems that bwm has raised
16:29:34 [jang_scri]
bwm: this test case runs in euler
16:29:37 [jang_scri]
not in cwm
16:29:46 [jang_scri]
danbri: so cwm isn't a complete reasoner.
16:29:51 [danbri]
16:29:54 [jang_scri]
ok, that test case stands as is.
16:29:59 [jang_scri]
16:30:57 [jang_scri]
ACTION just so I don't forget: turn test case names into frag IDs.
16:32:27 [jang_scri]
meeting closes
16:32:29 [danbri]
16:34:15 [danbri]
danbri has left #rdfcore
16:34:24 [Zakim]
16:34:45 [Zakim]
16:34:46 [Zakim]
16:34:46 [Zakim]
16:34:50 [Zakim]
16:34:53 [Zakim]
16:34:54 [Zakim]
16:34:57 [Zakim]
16:34:58 [Zakim]
16:35:04 [Zakim]
16:35:05 [Zakim]
SW_RDFCore()10:00AM has ended
18:32:18 [AaronSw]
hm, what's rdfs:format?
18:32:56 [DanCon]
odd; google doesn't help find it...
18:33:03 [AaronSw]
18:33:32 [DanCon]
a low-impact datatypes proposal: rdfs:format
18:34:01 [AaronSw]
ooh, that would be a dream
18:34:25 [AaronSw]
i could even cover it in the primer primer
18:37:59 [DanCon]
I'm trying to figure out what's the best time/place/manner to reopen the datatypes question
18:41:12 [AaronSw]
maybe collect support offline and reopen it if it seems workable?
18:42:03 [DanCon]
something like that.
18:42:40 [AaronSw]
well, i support rdfs:format followed by the null solution
19:02:19 [Zakim]
Zakim has left #rdfcore