IRC log of rdf-wg on 2011-05-25

Timestamps are in UTC.

14:32:52 [RRSAgent]
RRSAgent has joined #rdf-wg
14:32:52 [RRSAgent]
logging to
14:32:54 [trackbot]
RRSAgent, make logs world
14:32:54 [Zakim]
Zakim has joined #rdf-wg
14:32:56 [trackbot]
Zakim, this will be 73394
14:32:56 [Zakim]
ok, trackbot; I see SW_RDFWG()11:00AM scheduled to start in 28 minutes
14:32:57 [trackbot]
Meeting: RDF Working Group Teleconference
14:32:57 [trackbot]
Date: 25 May 2011
14:33:10 [ivan]
Chair: David Wood
14:35:31 [AZ]
AZ has joined #rdf-wg
14:46:38 [Scott_Bauer]
Scott_Bauer has joined #rdf-wg
14:48:07 [FabGandon]
FabGandon has joined #rdf-wg
14:53:12 [cmatheus]
cmatheus has joined #rdf-wg
14:53:21 [Zakim]
SW_RDFWG()11:00AM has now started
14:53:28 [Zakim]
14:54:33 [FabGandon]
Zakim says code is not valid ?
14:54:49 [ivan]
14:54:52 [ivan]
may be early...
14:55:04 [ivan]
zakim, dial ivan-voip
14:55:04 [Zakim]
ok, ivan; the call is being made
14:55:05 [Zakim]
14:55:05 [Zakim]
14:55:06 [Zakim]
14:55:31 [Zakim]
14:55:46 [moustaki]
moustaki has joined #rdf-wg
14:55:54 [moustaki]
Zakim, who is on the phone?
14:55:54 [Zakim]
On the phone I see ??P6, Ivan, Tony
14:55:57 [Scott_Bauer]
zakim, Tony is me
14:55:57 [Zakim]
+Scott_Bauer; got it
14:56:02 [moustaki]
Zakim, ??P6 is yvesr
14:56:02 [Zakim]
+yvesr; got it
14:57:12 [Zakim]
14:57:22 [NickH]
Zakim, ??P9 is me
14:57:22 [Zakim]
14:57:22 [Zakim]
+NickH; got it
14:57:53 [AZ]
AZ has joined #rdf-wg
14:58:21 [mbrunati]
mbrunati has joined #rdf-wg
14:59:24 [Zakim]
14:59:36 [Zakim]
14:59:37 [AZ]
zakim, I am wcandillon
14:59:37 [Zakim]
ok, AZ, I now associate you with wcandillon
14:59:50 [mischat_]
mischat_ has joined #rdf-wg
15:00:02 [SteveH__]
SteveH__ has joined #rdf-wg
15:00:24 [cygri]
cygri has joined #rdf-wg
15:00:26 [zwu2]
zwu2 has joined #rdf-wg
15:00:29 [AlexHall]
AlexHall has joined #rdf-wg
15:00:32 [zwu2]
15:00:33 [NickH]
FabGandon, it works for me
15:00:35 [Zakim]
15:00:42 [zwu2]
what is the conference code?
15:00:45 [mbrunati]
zakim, P11 is me
15:00:45 [Zakim]
sorry, mbrunati, I do not recognize a party named 'P11'
15:00:54 [moustaki]
15:00:55 [Zakim]
15:01:00 [mbrunati]
zakim, ??P11 is me
15:01:00 [Zakim]
+mbrunati; got it
15:01:01 [zwu2]
15:01:06 [moustaki]
15:01:11 [SteveH]
Zakim, ??p12 is [Garlik]
15:01:12 [Zakim]
+[Garlik]; got it
15:01:12 [Zakim]
15:01:15 [zwu2]
zakim, what is the conference code?
15:01:15 [Zakim]
the conference code is 73394 (tel:+1.617.761.6200 tel:+ tel:+44.203.318.0479), zwu2
15:01:19 [SteveH]
Zakim, [Garlik] has SteveH, mischat
15:01:19 [Zakim]
+SteveH, mischat; got it
15:01:21 [Zakim]
15:01:33 [Zakim]
15:01:40 [ivan]
zakim, koalie is fabien
15:01:40 [Zakim]
+fabien; got it
15:01:46 [Zakim]
+ +1.415.586.aaaa
15:01:48 [Zakim]
15:01:51 [pfps]
pfps has joined #rdf-wg
15:01:53 [AndyS]
zakim, IPCaller is me
15:01:53 [Zakim]
+AndyS; got it
15:01:55 [Zakim]
15:01:58 [cmatheus]
zakim, ??P16 is me
15:01:58 [Zakim]
+cmatheus; got it
15:02:01 [FabGandon]
zakim, fabien is FabGandon
15:02:01 [Zakim]
+FabGandon; got it
15:02:15 [FabGandon]
Scribe: FabGandon
15:02:21 [Zakim]
15:02:22 [cygri]
zakim, mhausenblas is temporarily me
15:02:22 [Zakim]
+cygri; got it
15:02:33 [Zakim]
15:02:40 [cmatheus]
I'm muted on this end.
15:02:58 [ericP]
Zakim, please mute cmatheus
15:02:58 [Zakim]
cmatheus should now be muted
15:03:00 [Zakim]
15:03:03 [ericP]
Zakim, please unmute cmatheus
15:03:03 [Zakim]
cmatheus should no longer be muted
15:03:21 [Zakim]
15:03:29 [Zakim]
15:03:45 [AndyS]
zakim, ??P27 is me
15:03:45 [Zakim]
+AndyS; got it
15:03:47 [Souri]
Souri has joined #rdf-wg
15:03:50 [PatH]
PatH has joined #rdf-wg
15:04:14 [Zakim]
15:04:38 [cmatheus]
15:04:46 [ericP]
ack me
15:04:51 [cmatheus]
ack me
15:04:54 [ericP]
Zakim, unmute cmatheus
15:04:54 [Zakim]
cmatheus was not muted, ericP
15:05:11 [cmatheus]
let me hang up and try gain.
15:05:19 [Zakim]
15:05:20 [JeremyCarroll]
JeremyCarroll has joined #rdf-wg
15:05:32 [davidwood]
15:05:33 [Zakim]
15:05:36 [davidwood]
On my way now
15:05:37 [davidwood]
15:05:53 [ww]
ww has joined #rdf-wg
15:05:55 [cmatheus]
zakim, ??P16 is me
15:05:55 [Zakim]
+cmatheus; got it
15:06:24 [davidwood]
Err, I can't join the telecon; "The conference is restricted at this time"
15:06:34 [sandro]
sure you got the right code...??
15:06:42 [davidwood]
Sandro, yes
15:06:51 [SteveH]
maybe up to our limit?
15:06:52 [sandro]
(that's the error if you make a typo in the conference code.)
15:06:57 [Zakim]
15:07:01 [ww]
zakim, ??P31 is me
15:07:01 [Zakim]
+ww; got it
15:07:11 [ericP]
davidwood, if that fails, i can try to conference you in with my phone
15:07:12 [ww]
zakim, mute me
15:07:12 [Zakim]
ww should now be muted
15:07:48 [davidwood]
No, still getting the error message :(
15:07:59 [Zakim]
15:08:04 [Zakim]
15:08:23 [davidwood]
EricP, thanks. I'm at +1.540.898.1842
15:08:51 [sandro]
davidwood, the problem is likely in DTMF signalling between your phone and zakim -- some digit is probably not being conveyed correctly.
15:09:15 [ww]
rfc2833 ftw!
15:09:20 [davidwood]
Sandro, OK.
15:09:21 [sandro]
(I had a cell phone where 2/5/8/0 didn't make it through to Zakim.)
15:09:25 [davidwood]
Yay. Thanks to EricP.
15:10:13 [FabGandon]
15:11:08 [FabGandon]
davidwood: issues with missing parts in minutes from the last telecon
15:11:31 [FabGandon]
... no objection now to accept the minutes
15:11:41 [ivan]
15:11:51 [FabGandon]
RESOLVED: minutes from last meeting accepted
15:11:56 [davidwood]
ack ivan
15:11:57 [ivan]
ack ivan
15:12:36 [cygri]
15:12:45 [SteveH]
15:12:46 [FabGandon]
ivan: We should now try to register what ever name we chose.
15:12:51 [cygri]
15:12:52 [cygri]
15:12:57 [cygri]
steveh i'll let you go first
15:13:18 [FabGandon]
sandro: we need a formal document
15:13:26 [sandro]
15:13:33 [sandro]
... for IETF registration
15:14:37 [FabGandon]
ACTION: sandro to prepare document for IETF registration
15:15:15 [sandro]
ACTION: sandro to start conversation on reservince our well-known string.
15:15:15 [trackbot]
Created ACTION-52 - Start conversation on reservince our well-known string. [on Sandro Hawke - due 2011-06-01].
15:15:26 [davidwood]
ack cygri
15:15:39 [FabGandon]
davidwood: how long for the document to be ready?
15:15:57 [AZ]
AZ has joined #rdf-wg
15:16:09 [FabGandon]
cygri: we haven't decided on the name
15:16:20 [davidwood]
I expect the RDF Concepts doc to be in draft in the next couple of months.
15:16:22 [davidwood]
ack SteveH
15:16:53 [sandro]
+1 a WG Note on this
15:17:02 [yvesr]
15:17:02 [FabGandon]
SteveH: we could do a strow poll for the name now.
15:17:22 [SteveH]
genid for name
15:17:24 [cygri]
bnode, skolem, gensym, genid
15:17:28 [yvesr]
node for name?
15:17:34 [mischat]
15:17:35 [FabGandon]
davidwood: good ideas for a name?
15:17:39 [Zakim]
15:17:40 [yvesr]
15:17:40 [pfps]
I like steveH ;_)
15:17:47 [zwu2]
bnode, genid
15:18:09 [Zakim]
15:18:13 [AndyS]
I prefer a name that is for blank nodes - not generic (e.g. genid) that makes spotting bnodes harder.
15:18:16 [PatH]
I dont think we should use 'bnode'
15:18:19 [davidwood]
genid is good for me because it is short. well-known is not.
15:18:22 [SteveH]
+1 to PatH
15:18:30 [davidwood]
-1 to bnode
15:18:34 [JeremyCarroll]
I prefer something generic
15:18:38 [NickH]
15:18:42 [FabGandon]
ivan: a number of people against "bnode"
15:18:45 [JeremyCarroll]
-1 to bnode
15:19:05 [AZ]
+1 genid
15:19:10 [cygri]
+1 to skolem
15:19:16 [PatH]
I like genid as being the least bad of the lot.
15:19:18 [ivan]
+1 to skolem or genid
15:19:20 [AndyS]
Let's still do the strawpoll to get the balance.
15:19:21 [SteveH]
+1 t genid
15:19:22 [davidwood]
skolem would work for me
15:19:27 [FabGandon]
+1 genid
15:19:30 [yvesr]
-1 to skolem
15:19:32 [mbrunati]
probably blank or skolem
15:19:43 [yvesr]
we should remove that whole skolem terminology imho
15:19:50 [ww]
and blank?
15:19:54 [zwu2]
no -1 to genid yet
15:20:02 [FabGandon]
davidwood: strow poll bnode, genid and skolem
15:20:04 [davidwood]
Straw poll: bnode, genid or skolem
15:20:11 [sandro]
STRAWPOLL: bnode, genid, skolem --- numbers of each
15:20:22 [SteveH]
-1, +1, -1
15:20:22 [AndyS]
The argument was for other systems to be able spot these from the URI.
15:20:23 [PatH]
15:20:25 [sandro]
+0 / +1 / +0
15:20:25 [davidwood]
+1 genid, +1 skolem, −1 bnode
15:20:27 [ivan]
15:20:30 [ww]
+2, +1 +3
15:20:30 [zwu2]
15:20:31 [yvesr]
+1, 0, -1
15:20:35 [NickH]
-1, +1, -1
15:20:41 [cygri]
15:20:43 [PatH]
You did say 'slash'
15:20:50 [mbrunati]
-1, , -1, 0
15:20:53 [AlexHall]
15:20:53 [AndyS]
15:20:58 [JeremyCarroll]
15:21:05 [mischat]
15:21:05 [ww]
15:21:08 [Souri]
15:21:10 [ww]
15:21:23 [ww]
zakim, unmute me
15:21:23 [Zakim]
ww should no longer be muted
15:21:43 [cygri]
15:21:44 [mbrunati]
mmm, skolem +1
15:21:51 [sandro]
ww: it's entirely aesthetics and personal preference -- nothing serious -- just a name.
15:21:55 [FabGandon]
ww: no serious issues with any of the names
15:22:06 [ww]
zakim, mute me
15:22:06 [Zakim]
ww should now be muted
15:22:18 [FabGandon]
mbrunati: why not use blank nodes?
15:22:21 [davidwood]
ack cygri
15:22:44 [ww]
15:22:48 [SteveH]
15:22:57 [FabGandon]
cygri: the text of the resolution introduces the term "skolem"
15:23:03 [yvesr]
15:23:04 [AndyS]
It stands in place of a blank node and it is distinguished by this. so my "skolem" -1 => 0
15:23:13 [davidwood]
ack SteveH
15:23:15 [FabGandon]
... not too much sense to discuss that now
15:23:47 [cygri]
15:23:54 [AZ]
AZ has joined #rdf-wg
15:23:59 [PatH]
They are skolem constants but the 'skolem' terminiology goes way beyond this usage, so this is a very simple case.
15:24:01 [FabGandon]
SteveH: introduce more obscure jargon is not a good idea and the term "genid" is already largely used
15:24:02 [davidwood]
ack yvesr
15:24:02 [NickH]
I agree with SteveH
15:24:32 [sandro]
but which is jargon?? "skolem" is in every dictionary.
15:24:42 [FabGandon]
yvesr: we have enough jargon to not include yet another term
15:24:58 [davidwood]
ack cygri
15:25:08 [FabGandon]
davidwood: strow poll favours genid and objections have been resolved
15:26:13 [JeremyCarroll]
15:26:15 [PatH]
Welcome to a working group :-)
15:26:19 [FabGandon]
cygri: we need a real formal term for thisand on the list skolem was the last candidates
15:26:22 [davidwood]
ack JeremyCarroll
15:27:02 [FabGandon]
JeremyCarroll: we are talking about the scheme name not the text.
15:27:17 [FabGandon]
davidwood: strow poll to accept "genid"
15:27:18 [sandro]
PROPOSED: We'll use genid as the .well-known name
15:27:21 [ericP]
15:27:21 [SteveH]
15:27:23 [AZ]
15:27:25 [yvesr]
15:27:26 [zwu2]
15:27:27 [JeremyCarroll]
15:27:27 [mbrunati]
15:27:28 [AndyS]
Abstain (a +0)
15:27:28 [cmatheus]
15:27:29 [FabGandon]
15:27:29 [sandro]
15:27:30 [ww]
15:27:30 [davidwood]
15:27:30 [cygri]
15:27:30 [ivan]
15:27:35 [mischat]
15:27:37 [pfps]
15:27:38 [NickH]
15:27:39 [PatH]
15:27:48 [sandro]
RESOLVED: We'll use genid as the .well-known name
15:28:18 [SteveH]
are we having a note on genid?
15:28:33 [sandro]
close action-49
15:28:34 [trackbot]
ACTION-49 Propose revised wording for ISSUE-40 text, possibly also for section 3.2 of RDF Concepts closed
15:28:37 [sandro]
close action-50
15:28:37 [trackbot]
ACTION-50 Propose edits to fix wording for ISSUE-40 resolution closed
15:28:56 [ww]
second proposal for note on genid
15:28:58 [AndyS]
+1 to a note or some such WG communication (more than a blog entry)
15:29:14 [SteveH]
I can contribute to a note if we want one
15:29:50 [FabGandon]
cygri: I can update the editor draft on my machine and put it online.
15:29:54 [SteveH]
15:30:12 [FabGandon]
sandro: we need a note or Rec publish I think
15:30:42 [FabGandon]
... are we going to do the note or note?
15:31:07 [FabGandon]
davidwood: we can then republish content in the concept doc.
15:31:19 [SteveH]
15:31:33 [FabGandon]
cygri: two paragraphs, not worth a note.
15:31:36 [SteveH]
I don't agree about visbility FWIW
15:31:43 [davidwood]
15:31:51 [FabGandon]
davidwood: do we agree not to produce a note?
15:32:02 [FabGandon]
... yes
15:32:25 [FabGandon]
15:32:36 [FabGandon]
15:33:05 [FabGandon]
15:33:13 [FabGandon]
PatH: done
15:33:47 [FabGandon]
... text in the wiki shows it's done
15:33:56 [FabGandon]
15:34:34 [FabGandon]
Topic: editing documents.
15:35:34 [FabGandon]
davidwood: email thread about editing documents
15:35:59 [FabGandon]
ivan: ready to create a repository / archive
15:36:28 [FabGandon]
... decide on short names for the docs
15:36:50 [FabGandon]
... decide on RDF 1.1 vs. RDF Next
15:37:24 [FabGandon]
... who should be on the front page of docs and what are the rules to decide the names that appear
15:38:34 [JeremyCarroll]
I have an AOB item - I will add myself to queue at end of meeting
15:38:56 [davidwood]
JeremyCarroll, ok
15:39:12 [FabGandon]
ivan: we should make it clear that we are not making a radical change to RDF
15:39:14 [SteveH]
I prefer year'd versions
15:39:25 [SteveH]
the current one tends to be called RDF '04 anyway
15:39:30 [sandro]
+1 years
15:39:41 [JeremyCarroll]
+1 to Ivan
15:39:42 [SteveH]
15:40:05 [cygri]
+1 to ivan ... but RDF '13 is also not bad
15:40:12 [davidwood]
ack SteveH
15:40:15 [FabGandon]
ivan: I prefer 1.1 since it reflects the additions we are making
15:40:47 [JeremyCarroll]
q+ to explain 2004
15:40:59 [FabGandon]
SteveH: 1.1 is the way forward if we go for numbers but I prefer the years.
15:41:01 [davidwood]
ack JeremyCarroll
15:41:01 [Zakim]
JeremyCarroll, you wanted to explain 2004
15:41:33 [sandro]
JeremyCarroll: "2004" results from that group having messed up, making bigger changes than they meant to.
15:41:46 [sandro]
JeremyCarroll: We should have given ourselves a number.
15:42:08 [FabGandon]
JeremyCarroll: having a number now is a good idea
15:42:14 [AndyS]
Is the rdf: URI changing ? (no) It has a year as has RDFS so year is a bit confusing here.
15:42:29 [SteveH]
AndyS, good point
15:42:33 [Zakim]
15:42:41 [pfps]
+1.1 :-)
15:42:46 [FabGandon]
JeremyCarroll: I like 1.1
15:43:03 [SteveH]
15:43:09 [AZ]
AZ has joined #rdf-wg
15:43:13 [mbrunati]
like 1.1
15:43:14 [Souri]
+1.1 to 1.1
15:43:22 [pfps]
+2 to 1.1
15:43:23 [AZ]
+1 to 1.1
15:43:24 [ericP]
15:43:26 [SteveH]
15:43:29 [cygri]
15:43:29 [zwu2]
+1 to 1.1
15:43:31 [AndyS]
15:43:31 [davidwood]
+1.1 to 1.1
15:43:31 [ivan]
15:43:32 [JeremyCarroll]
+1 to 1.1
15:43:33 [yvesr]
+1 to 1.1
15:43:34 [mischat]
15:43:38 [NickH]
15:43:39 [AlexHall]
15:43:41 [FabGandon]
davidwood: strow poll on 1.1 vs years
15:43:47 [FabGandon]
15:43:53 [zwu2]
is there a big difference between 1.1 or 1.5?
15:43:55 [ericP]
15:43:56 [ww]
1 + epsilon
15:44:36 [NickH]
I hope that there won't be enough changes for it to be 1.5
15:44:42 [ivan]
15:44:48 [sandro]
RESOLVED: The thing we're working on is "RDF 1.1"
15:44:48 [davidwood]
ack ivan
15:45:32 [SteveH]
SPARQL-WD added "11" to the end of the shortname
15:45:34 [FabGandon]
ivan: for the URIs all docs we produce have a short name + 1.1
15:46:48 [ivan]
15:47:13 [LeeF]
LeeF has joined #rdf-wg
15:47:34 [ivan]
15:47:44 [sandro]
vs rdf11-mt ?
15:47:57 [cygri]
+1 for rdf11-xxx
15:48:02 [FabGandon]
ivan: reuse of the old shot names should be considered only at the end when we publish the final rec
15:48:09 [SteveH]
+1 for rdf11-*
15:48:11 [AndyS]
actually it's /sparql11-update/ etc for SPARQL
15:49:30 [ivan]
PROPOSE: for short names we would use rdf11-X where rdf-X stands for the current recommendations
15:49:40 [yvesr]
15:49:41 [sandro]
15:49:42 [zwu2]
15:49:43 [SteveH]
15:49:44 [ericP]
15:49:44 [davidwood]
15:49:45 [ivan]
15:49:45 [cygri]
15:49:46 [mbrunati]
15:49:47 [AndyS]
15:49:48 [ww]
15:49:52 [AZ]
15:50:00 [NickH]
15:50:07 [Souri]
15:50:09 [pfps]
15:50:11 [PatH]
15:50:18 [FabGandon]
15:50:19 [PatH]
Hey, I *am* retired.
15:50:19 [ivan]
RESOLVED: for short names we would use rdf11-X where rdf-X stands for the current recommendations
15:50:41 [ivan]
15:50:41 [trackbot]
ISSUE-8 -- Incorporate IRI-s into the RDF documents -- open
15:50:41 [trackbot]
15:50:56 [FabGandon]
Topic: Incorporate IRI-s into the RDF documents
15:51:14 [JeremyCarroll]
15:51:22 [AndyS]
15:52:02 [ericP]
q+ to say there are a zillion forms of protocol dependent and independent normalizations which we don't want to do
15:52:12 [pfps]
+1 to ericP
15:52:18 [SteveH]
q+ to talk about security
15:52:23 [AlexHall]
15:52:24 [davidwood]
ack ericP
15:52:24 [Zakim]
ericP, you wanted to say there are a zillion forms of protocol dependent and independent normalizations which we don't want to do
15:52:25 [AndyS]
Unless we go for completely normalising IRIs e.g. /x/../y (which i bleive is wrong by IRI spec as RDF does not produce the IRI - it transfers it)
15:52:27 [ericP]
various unicode forms etc.
15:52:45 [ivan]
-> mail including Felix Sasaki's advice
15:52:58 [davidwood]
ack SteveH
15:52:58 [Zakim]
SteveH, you wanted to talk about security
15:53:24 [PatH]
15:53:35 [FabGandon]
SteveH: we should keep in mine that RDF is for machines
15:53:43 [JeremyCarroll]
Note: Applications using IRIs as identity tokens with no relationship to a protocol MUST use the Simple String Comparison (see section 5.3.1).
15:53:48 [FabGandon]
15:53:51 [JeremyCarroll]
from RFC 3987
15:54:01 [ivan]
PROPOSE: resolve ISSUE 8 by keeping IRI-s and their punycode equivalent separate as different URI References in RDF
15:54:02 [JeremyCarroll]
this decision is in accord with this REQUIREMENT
15:54:16 [ericP]
15:54:17 [SteveH]
15:54:17 [JeremyCarroll]
15:54:18 [AlexHall] --> proposed text from ericP which defines IRI equivalence as Unicode equivalence and discourages use of punycode in RDF IRIs
15:54:18 [zwu2]
15:54:19 [PatH]
Do we call them URI refs or IRIs?
15:54:19 [ww]
15:54:20 [ivan]
15:54:23 [pfps]
15:54:24 [mbrunati]
15:54:25 [FabGandon]
15:54:27 [AlexHall]
15:54:30 [PatH]
+1 to substance
15:54:46 [AndyS]
15:54:50 [PatH]
15:55:00 [Souri]
15:55:19 [FabGandon]
ivan: we use URI Ref and then editors can change it to IRIs
15:55:19 [ivan]
RESOLVED: resolve ISSUE 8 by keeping IRI-s and their punycode equivalent separate as different URI References in RDF
15:55:39 [AndyS]
caution: URI reference != RDF URI reference
15:55:53 [sandro]
ACTION: david to officially inform the RDFa WG of our decision on ISSUE-8
15:55:53 [trackbot]
Created ACTION-53 - Officially inform the RDFa WG of our decision on ISSUE-8 [on David Wood - due 2011-06-01].
15:56:18 [JeremyCarroll] link for my quote
15:56:41 [ivan]
15:56:41 [trackbot]
ISSUE-12 -- Reconcile various forms of string literals (time permitting) -- open
15:56:41 [trackbot]
15:56:44 [FabGandon]
15:56:44 [trackbot]
ISSUE-12 -- Reconcile various forms of string literals (time permitting) -- open
15:56:44 [trackbot]
15:57:13 [cygri]
15:57:13 [davidwood]
15:57:19 [FabGandon]
Topic: Reconcile various forms of string literals
15:58:43 [sandro]
scribe: sandro
15:59:05 [AndyS]
I asked Jena users and developers for feedback - little enthusiasm - more energy for just xsd:string to be deprecated and no other changes made.
15:59:07 [sandro]
davidwood: Looking at the mailing list, it seems possible to get a decision.
15:59:14 [sandro]
15:59:24 [FabGandon]
FabGandon has left #rdf-wg
15:59:25 [PatH]
There was some email pushback...
15:59:31 [SteveH]
I'm not exactly happy about the proposal
15:59:36 [davidwood]
ack sandro
15:59:40 [pfps]
I'm not happy with the proposal.
15:59:43 [SteveH]
15:59:45 [Zakim]
15:59:53 [cygri]
16:00:34 [SteveH]
+1 to sandro, lang dt equivalent is cleaning, from some perspective
16:00:41 [SteveH]
16:00:48 [JeremyCarroll]
q+ to accept action to respond to this qu in e mail
16:00:53 [davidwood]
ack SteveH
16:01:02 [sandro]
sandro: what happened to the proposal to use languages as datatypes?
16:01:20 [pfps]
16:01:28 [JeremyCarroll]
+1 to steve
16:01:30 [sandro]
steve: I don't find this idea of non-string lexical values as a win. it's a step sideways.
16:01:36 [ww]
I think @en is (should be) syntax sugar for ^^englishString
16:02:04 [sandro]
PatH: Yes, it's a step sideways. Motivated by trying to preserve the current situation. Not a magic solution, preserve status quo.
16:02:19 [davidwood]
Might @en become a subclass of xsd:string?
16:02:21 [sandro]
SteveH: Why take lanugage tags off the table? I found that fairly elegant.
16:02:42 [AndyS]
16:02:53 [sandro]
PatH: Can you make that more preceise? Making language tags, with their complex subtyping, into datatypes would be hell.
16:02:58 [ivan]
-> as a good reference on language tags
16:03:08 [cygri]
16:03:17 [davidwood]
ack cygri
16:03:18 [sandro]
SteveH: Some URI prefix then all possible language tags -- that would be kind of ugly, but all proposals here are ugly.
16:03:20 [ivan]
This is the generic form of a language tag: language-extlang-script-region-variant-extension-privateuse
16:03:27 [davidwood]
ack JeremyCarroll
16:03:27 [Zakim]
JeremyCarroll, you wanted to accept action to respond to this qu in e mail
16:03:28 [SteveH]
we'd end up with rdflang:<all-possible-langtags>
16:03:33 [AndyS]
q+ to say that we don't need to do the subtyping / lang tag meaning (which I agree is a bad fit)
16:03:37 [PatH]
Sorry to jump in.
16:03:44 [PatH]
+1 Andy
16:03:57 [AndyS]
16:04:05 [sandro]
JeremyCarroll: Language tags are complicated, they are NOT like data types, but it's too complex to explain why they are not. Like chinese -- very complex language tags. Cannot be discussed on the call.
16:04:29 [ww]
I would like an explanation
16:04:31 [ivan]
zh-Hant-HK is the chinese with traditional scripts as used in Hong Kong
16:04:32 [PatH]
Suggest anyone who likes the idea, go read the lang tag spec.
16:04:34 [SteveH]
ASK { rdflang:en ?p rdflang:en-GB } => false
16:04:43 [ww]
(on the list)
16:04:47 [sandro]
sandro: I disagree with some of your argument --- a string is a string.
16:04:55 [davidwood]
ack pfps
16:05:01 [sandro]
sandro: it is ENTIRELY IRRELEVANT how complex the matching is.
16:05:03 [PatH]
No, they are NOT strings. Tags are related in complex ways which do not correspond to simple language/dialect or language/region cases.
16:05:39 [sandro]
pfps: I;m very uncomforable changing the guts of RDF for minimal benefit
16:05:48 [PatH]
Peter, can you point to an actual problem? I can see none.
16:05:59 [JeremyCarroll]
I agree with peter too!!
16:06:04 [sandro]
pfps: I'm against the proposal overall -- not specifically agreeing with JeremyCarroll (although I do, also)
16:06:21 [ivan]
16:06:22 [sandro]
davidwood: i think it's clear we wont resolve this today.
16:06:26 [pfps]
I didn't say that there was a problem, just that there was a change to the fundamentals of RDF.
16:06:35 [sandro]
davidwood: JeremyCarroll is right to move it to the list.
16:06:43 [PatH]
So, its our task to adjust RDF to suit ourselves.
16:06:45 [JeremyCarroll]
my AOB: Propose congratulations to Gavin, Kindli and Patrick
16:07:03 [sandro]
topic: Other Business
16:07:21 [PatH]
16:07:30 [sandro]
JeremyCarroll: Gavin and Kindli had a baby, Patrick (1 day old). :-)
16:07:39 [PatH]
Particularly as he has a really good name.
16:07:48 [zwu2]
16:08:52 [sandro]
sandro: how abuut we talk about the issue for remaining time...
16:08:57 [sandro]
topic: back to String Literals
16:09:11 [sandro]
pfps: I don't see what motivates all this in RDF.
16:09:18 [ww]
language tags seem like a funny special case in the data model
16:09:22 [PatH]
The change is mathematically trivial, and the RDF machinery isnt very deep anyway.
16:09:27 [ericP]
the choice between "abc" and "abc"^^xsd:string is arbitrary and stymies unification
16:09:35 [ericP]
16:10:18 [PatH]
We have to choose between tyuping a <striong, tag> pair, or writing "string@tag"
16:10:28 [sandro]
pfps: Conceptually, adding a slew of datatypes (eg rdflang:en) that doesnt change the machinery nearly as much. I don't think it's a good idea, but it's not really a chance to the RDF machinery.
16:10:32 [AZ]
AZ has joined #rdf-wg
16:10:39 [sandro]
pfps: is there an infinite number of languages?
16:10:45 [sandro]
ivan: it's large but finite.
16:10:59 [ivan]
16:11:33 [ww]
zakim, unmute me
16:11:33 [Zakim]
ww should no longer be muted
16:12:03 [davidwood]
16:12:10 [ivan]
16:12:13 [sandro]
PatH: infinity is not the problem -- it's the complexity.
16:12:14 [sandro]
16:12:14 [ww]
16:12:18 [JeremyCarroll]
one human being can invent an infinite number of privater use tags
16:12:29 [ivan]
16:12:31 [sandro]
pat: region means something differrent in each language.
16:12:33 [davidwood]
ack ivan
16:13:14 [davidwood]
ack ww
16:13:14 [sandro]
ww: i agree it's very complicated; one might want to model this stuff in RDF -- if it's a datatype, then there's a chance of doing that. This simplifies this.
16:13:18 [AndyS]
and lang tag canonicalization is quite complex (it's not lower case)
16:13:25 [ericP]
16:13:26 [sandro]
16:13:30 [sandro]
ack sandro
16:13:31 [davidwood]
ack sandro
16:13:43 [PatH]
Language tags are widely used. Has any of these uers expressed a desire to replace them with RDF?
16:13:48 [SteveH]
AndyS, there is a (RDF?) document that recommends lowercase normalisation
16:13:55 [SteveH]
AndyS, abstract syntax maybe
16:14:03 [ivan]
16:14:06 [davidwood]
q+ to ask about language tags and xsd:string
16:14:07 [PatH]
16:14:10 [SteveH]
16:14:11 [ww]
zakim, mute me
16:14:12 [davidwood]
ack ericP
16:14:13 [Zakim]
ww should now be muted
16:14:17 [sandro]
sandro: just treat the language tag as opaque. that's the most elegant solution here.
16:14:25 [AndyS]
steveH - yes. It's not what RFC 4646 says :-(
16:14:35 [PatH]
That does not work when datatype names become class names in RDFS.
16:14:57 [PatH]
We need to establish the subclass relationships.
16:15:00 [sandro]
ericP: The only logic I've seen on language tags is LangMatches, as in SPARQL -- simple to implement -- I think Sandro's notion of treat them as opaque, or a little bit if "L-Entailment" on lang-matches, those tricks would work and simplify RDF.
16:15:31 [davidwood]
ack ivan
16:15:35 [PatH]
What is the value space of these dataypes??
16:16:14 [sandro]
ivan: Because the number of lang dts is huge, I want to be clear that we do not introduce into an RDFS reasoner to put in a huge number of class definitions.
16:16:20 [sandro]
(of course not, if it's opaque.)
16:16:39 [sandro]
ivan: "every datatype we know has to be defined to be a class".
16:16:48 [ww]
I'm finding this discussion very interesting but unfortunately have to go
16:16:48 [sandro]
ivan: we can wave our hands, perhaps.
16:17:06 [AZ]
AZ has joined #rdf-wg
16:17:08 [AndyS]
PatH - exactly! - they have (presumably) the same value space (unicode string) which is a problem.
16:17:14 [sandro]
ivan: In OWL-2-RL every DT needs a triple added.
16:17:27 [Zakim]
16:17:37 [sandro]
I suggest value space is pairs (string, language tag).
16:17:38 [PatH]
The classes of valaues are determined by the datatype spec. If we follow the tag specs, this gets very com,plicated and might not even fit into the RDFS class model. If we don't, the our reasoners will not deal with the datatypes correctly.
16:17:58 [sandro]
16:18:03 [JeremyCarroll]
16:18:05 [Zakim]
16:18:06 [Zakim]
16:18:07 [Zakim]
16:18:09 [Zakim]
16:18:09 [zwu2]
16:18:09 [AZ]
16:18:10 [mbrunati]
ok, bye
16:18:10 [Zakim]
16:18:13 [Zakim]
16:18:14 [Zakim]
16:18:14 [Zakim]
16:18:15 [Zakim]
16:18:15 [Zakim]
16:18:20 [Zakim]
16:18:23 [Zakim]
16:18:29 [Zakim]
16:18:56 [AlexHall]
AlexHall has left #rdf-wg
16:19:12 [SteveH]
SteveH has joined #rdf-wg
16:19:23 [mischat]
mischat has joined #rdf-wg
16:20:04 [AndyS]
WGs have a habit of un-deciding over their lifetime.
16:26:29 [AndyS]
I disagree
16:26:49 [AndyS]
16:27:59 [SteveH]
16:31:09 [sandro]
zakim, who is on the call?
16:31:09 [Zakim]
On the phone I see Ivan, EricP, sandro, AndyS
16:33:24 [AndyS]
sandro :-)
16:35:40 [AndyS]
See the text
16:39:23 [ivan]
andy, is this the reference I should use for the rdflib group?
16:39:34 [AndyS]
ivan - email coming
16:39:37 [ivan]
16:40:27 [AndyS]
ivan - email sent
16:44:34 [AndyS]
oddly, N3 does not even have \u (I can't find it anyway)
16:45:44 [Zakim]
16:45:46 [Zakim]
16:45:48 [Zakim]
16:45:49 [Zakim]
16:45:49 [Zakim]
SW_RDFWG()11:00AM has ended
16:45:51 [Zakim]
Attendees were Ivan, Scott_Bauer, yvesr, EricP, NickH, wcandillon, Peter_Patel-Schneider, mbrunati, SteveH, mischat, AlexHall, +1.415.586.aaaa, AndyS, zwu2, cmatheus, FabGandon,
16:45:54 [Zakim]
... cygri, sandro, PatH, Souri, JeremyCarroll, ww, nick
17:17:36 [SteveH__]
SteveH__ has joined #rdf-wg
18:12:51 [SteveH]
SteveH has joined #rdf-wg
18:13:03 [SteveH__]
SteveH__ has joined #rdf-wg
18:50:12 [Zakim]
Zakim has left #rdf-wg
19:15:48 [davidwood]
davidwood has joined #rdf-wg
19:46:41 [mischat]
mischat has joined #rdf-wg
21:04:59 [cygri]
cygri has joined #rdf-wg
21:38:49 [ww]
ww has left #rdf-wg
21:40:54 [mischat]
mischat has joined #rdf-wg
23:30:29 [davidwood]
davidwood has joined #rdf-wg
23:50:41 [LeeF]
LeeF has joined #rdf-wg