IRC log of sparql on 2010-11-30

Timestamps are in UTC.

14:54:50 [RRSAgent]
RRSAgent has joined #sparql
14:54:50 [RRSAgent]
logging to http://www.w3.org/2010/11/30-sparql-irc
14:54:52 [trackbot]
RRSAgent, make logs world
14:54:52 [Zakim]
Zakim has joined #sparql
14:54:54 [trackbot]
Zakim, this will be 77277
14:54:54 [Zakim]
ok, trackbot; I see SW_(SPARQL)10:00AM scheduled to start in 6 minutes
14:54:55 [trackbot]
Meeting: SPARQL Working Group Teleconference
14:54:55 [trackbot]
Date: 30 November 2010
14:54:56 [LeeF]
zakim, this will be SPARQL
14:54:56 [Zakim]
ok, LeeF; I see SW_(SPARQL)10:00AM scheduled to start in 6 minutes
14:55:08 [Zakim]
SW_(SPARQL)10:00AM has now started
14:55:15 [Zakim]
+??P6
14:55:25 [cbuilara]
Zakim, +??p6 is me
14:55:26 [Zakim]
sorry, cbuilara, I do not recognize a party named '+??p6'
14:55:38 [cbuilara]
Zakim, ??P6
14:55:38 [Zakim]
I don't understand '??P6', cbuilara
14:55:48 [cbuilara]
Zakim, ??P6 is me
14:55:48 [Zakim]
+cbuilara; got it
14:55:54 [Zakim]
+ +1.617.553.aaaa
14:56:02 [LeeF]
zakim, aaaa is me
14:56:02 [Zakim]
+LeeF; got it
14:56:05 [NicoM]
NicoM has joined #sparql
14:56:28 [Zakim]
+??P9
14:56:35 [AndyS]
zakim, ??P9 is me
14:56:35 [Zakim]
+AndyS; got it
14:57:12 [Zakim]
+kasei
14:57:38 [Zakim]
+??P11
14:58:01 [NickH]
zakim, ??P11 is me
14:58:01 [Zakim]
+NickH; got it
14:58:49 [Zakim]
+ +49.911.973.4.aabb
14:59:04 [NicoM]
Zakim, +49 is me
14:59:04 [Zakim]
+NicoM; got it
14:59:09 [LeeF]
Chair: LeeF
14:59:24 [LeeF]
Regrets: SteveH
15:00:02 [LeeF]
Agenda: http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0316.html
15:00:07 [Zakim]
+OlivierCorby
15:00:21 [MattPerry]
MattPerry has joined #sparql
15:00:23 [kasei]
Zakim, who is talking?
15:00:34 [Zakim]
kasei, listening for 11 seconds I heard sound from the following: AndyS (17%)
15:00:56 [kasei]
very strage sounding jack hammers :)
15:01:04 [LeeF]
ok, guess not then :)
15:01:06 [bglimm]
bglimm has joined #sparql
15:01:11 [LeeF]
zakim, who's on the phone?
15:01:11 [Zakim]
On the phone I see cbuilara, LeeF, AndyS, kasei, NickH, NicoM, OlivierCorby
15:01:17 [Zakim]
+MattPerry
15:01:21 [LeeF]
Scribenick: cbuilara
15:01:23 [LeeF]
Scribe: Carlos
15:02:11 [cbuilara]
I can't hear anything
15:03:11 [LeeF]
topic: Admin
15:03:24 [LeeF]
PROPOSED: Approve minutes from http://www.w3.org/2009/sparql/meeting/2010-11-23
15:03:44 [Zakim]
+AxelPolleres
15:03:46 [cbuilara]
LeeF: shorcuts, ask for feedback
15:04:04 [cbuilara]
Leef: ask for implementation experience about update shorcuts
15:04:22 [LeeF]
RESOLVED: Approve minutes from http://www.w3.org/2009/sparql/meeting/2010-11-23
15:04:50 [LeeF]
Next meeting: Tuesday December 7, 2010 - usual time and place
15:04:58 [pgearon_]
pgearon_ has joined #sparql
15:05:02 [cbuilara]
LeeF: next meeting 7th December
15:05:16 [LeeF]
regrets next week: cbuilara
15:05:26 [OlivierCorby]
regrets also
15:05:43 [AlexPassant]
regrets for me next week as well
15:05:55 [cbuilara]
LeeF: 28th week of
15:06:21 [cbuilara]
LeeF: tomorrow working session on protocol document
15:06:27 [Zakim]
+ +44.186.528.aacc
15:06:27 [Zakim]
-AndyS
15:06:42 [Zakim]
+ +1.540.841.aadd
15:06:45 [Zakim]
+[IPcaller]
15:06:48 [cbuilara]
LeeF: goal: have a protocol proposal for the group
15:06:52 [Zakim]
+Sandro
15:06:55 [bglimm]
Zakim, +44.186.528.aacc is me
15:06:55 [Zakim]
+bglimm; got it
15:06:55 [AlexPassant]
Zakim: IPcaller is me
15:07:01 [pgearon_]
Zakim: aadd is me
15:07:05 [Zakim]
+??P31
15:07:06 [sandro]
(sorry I'm late)
15:07:11 [AlexPassant]
Zakim: +IPcaller is me
15:07:14 [cbuilara]
LeeF: look at the entailment decision for next week
15:07:17 [AndyS]
zakim, ??P31 is me
15:07:17 [Zakim]
+AndyS; got it
15:07:19 [AxelPolleres]
next week... entailment decisions
15:07:26 [LeeF]
zakim, who's on the phone?
15:07:26 [Zakim]
On the phone I see cbuilara, LeeF, kasei, NickH, NicoM, OlivierCorby, MattPerry, AxelPolleres, bglimm, +1.540.841.aadd, [IPcaller], Sandro, AndyS
15:07:36 [LeeF]
zakim, aadd is pgearon
15:07:36 [Zakim]
+pgearon; got it
15:07:38 [AlexPassant]
Zakim, [IPcaller] is me
15:07:38 [Zakim]
+AlexPassant; got it
15:08:30 [LeeF]
topic: which functions?
15:08:41 [Zakim]
-AlexPassant
15:09:12 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error and the DateTime functions
15:09:45 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error and the DateTime functions and including RAND() and RAND(X)
15:09:47 [Zakim]
+AlexPassant
15:10:38 [kasei]
q+
15:10:42 [LeeF]
ack kasei
15:10:55 [AndyS]
CONCAT has turned out *potentially* different.
15:11:18 [pgearon_]
q+
15:11:25 [cbuilara]
kasei: date time functions are useful, but not necessarily with keywords
15:11:31 [LeeF]
ack pgearon
15:11:52 [cbuilara]
pgearon: everybody is going to support date/time, I want see them in
15:12:43 [cbuilara]
AndyS: we need to provide keywords, testing to see where the boundary is here
15:13:19 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error and including RAND() and RAND(X)
15:13:50 [cbuilara]
LeeF: is anybody who feels that should not include these date time functions?
15:13:55 [LeeF]
(silence)
15:14:02 [NickH]
:-D
15:14:19 [LeeF]
fn:round-half-to-even
15:14:30 [cbuilara]
LeeF: dodgy ones are fn:round-half-to-even
15:14:34 [LeeF]
s/LeeF/AndyS
15:15:42 [cbuilara]
AndyS: dodgyness: in applications type float cannot be intuitive
15:15:49 [AndyS]
http://www.w3.org/TR/xpath-functiofn:round-half-to-evenns/#func-round-half-to-even
15:16:02 [AndyS]
http://www.w3.org/TR/xpath-functions/#func-round-half-to-even
15:16:08 [cbuilara]
LeeF: anybody has experinece with that round-half-to-even?
15:16:21 [cbuilara]
s/experinece/experience
15:16:34 [kasei]
was the proposal to implement both the one and two-arg forms?
15:17:28 [AxelPolleres]
anyone really wanting round-half-to-even?
15:17:35 [cbuilara]
LeeF: the proposal would be to include all forms, all related functions of fn form
15:18:11 [cbuilara]
AndyS: omitt it
15:18:13 [AxelPolleres]
+1 omit
15:18:16 [pgearon_]
+1 omit
15:18:23 [OlivierCorby]
+1 omit
15:18:26 [cbuilara]
LeeF: ample support for omit
15:18:31 [NickH]
+1 omit
15:18:36 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error, fn:round-half-to-even and including RAND() and RAND(X)
15:19:36 [ericP]
±1
15:19:48 [cbuilara]
AndyS: the operators are already there, the question is whether you can call it
15:20:02 [AxelPolleres]
http://www.w3.org/2009/sparql/docs/query-1.1/rq25.xml#OperatorMapping
15:20:49 [cbuilara]
ANdyS: the ones missing are dates ones
15:21:01 [cbuilara]
s/ANdyS/AndyS
15:21:10 [AxelPolleres]
q+
15:21:15 [LeeF]
ack AxelPolleres
15:21:41 [Zakim]
+Chimezie_Ogbuji
15:21:45 [cbuilara]
LeeF: leave it as it is, comparison support, add date time functions
15:21:50 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error, fn:round-half-to-even, the op:* and including RAND() and RAND(X)
15:22:29 [AndyS]
http://www.w3.org/TR/xpath-functions/#func-compare
15:22:43 [chimezie]
chimezie has joined #sparql
15:22:50 [cbuilara]
LeeF: I'd like to include it, opinions?
15:22:53 [AxelPolleres]
kind of a coinflip decision...
15:23:00 [chimezie]
Zakim, mute me
15:23:00 [Zakim]
Chimezie_Ogbuji should now be muted
15:23:16 [AxelPolleres]
the version without collation can be emulated with IF( ... < ...
15:23:36 [cbuilara]
LeeF: any implementors have an opinion
15:23:57 [pgearon_]
+q
15:24:12 [LeeF]
ack pgearon
15:24:40 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error, fn:round-half-to-even, the op:*, and fn:compare and including RAND() and RAND(X)
15:24:43 [cbuilara]
pgearon: I'd put it in, happy not to require to implementors
15:24:50 [pgearon_]
+q
15:24:56 [LeeF]
ack pgearon
15:25:33 [chimezie]
+1
15:25:34 [AlexPassant]
might be nice
15:25:48 [LeeF]
pgearon: what about a SHA1 function?
15:25:53 [cbuilara]
pgearon: wondering if SHA1 one should be part of the function library?
15:25:55 [AndyS]
http://commons.apache.org/codec/apidocs/org/apache/commons/codec/digest/DigestUtils.html
15:25:55 [AlexPassant]
md5 may be useful as well
15:26:05 [chimezie]
128 bit length?
15:26:14 [AlexPassant]
especially for UPDATE (e.g. inserting passwords in a rdf store)
15:26:17 [cbuilara]
LeeF: useful thing
15:26:25 [AlexPassant]
q+
15:26:29 [cbuilara]
LeeF: but not sure if we require it
15:26:32 [kasei]
LeeF, no.
15:26:52 [LeeF]
kasei: thinks it's useful, but shouldn't be required.
15:26:55 [AlexPassant]
q-
15:27:14 [LeeF]
AndyS: what would it return?
15:27:46 [kasei]
I'd imagine you'd get out of it the ability to query for foaf:mbox_sha1sum based on an foaf:mbox value.
15:28:24 [cbuilara]
LeeF: would anybody be unhappy including SHA1?
15:28:32 [AlexPassant]
sha1 only or also md5 ?
15:29:09 [cbuilara]
pgearon: send an email to the list with the proposal about SHA1 and md5
15:29:26 [LeeF]
PROPOSED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error, fn:round-half-to-even, the op:*, and fn:compare and including RAND() and RAND(X)
15:29:43 [Zakim]
-OlivierCorby
15:29:49 [cbuilara]
LeeF: any addition to this proposal? if there is ample support we will ammend the list
15:30:08 [AxelPolleres]
seconded
15:30:08 [AndyS]
seconded
15:30:13 [cbuilara]
LeeF: changes to this proposal?
15:30:23 [Zakim]
+OlivierCorby
15:30:26 [cbuilara]
LeeF: any abstention or reject?
15:30:28 [LeeF]
RESOLVED: SPARQL 1.1 Function Library includes all of the functions listed at http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0188.html except for fn:error, fn:round-half-to-even, the op:*, and fn:compare and including RAND() and RAND(X), no objections or abstentions
15:31:00 [LeeF]
topic: keywords and URIs
15:31:05 [cbuilara]
LeeF: next keywords and URIs
15:32:07 [cbuilara]
LeeF: do all of these functions get their keywords in the SPARQL syntaxt and if not can we get keywords? good to have keywords for all functions? opinions?
15:32:10 [AxelPolleres]
+1 for keyqords
15:32:13 [AxelPolleres]
q+
15:32:17 [LeeF]
ack AxelPolleres
15:32:38 [bglimm]
Zakim, mute me
15:32:38 [Zakim]
bglimm should now be muted
15:33:18 [cbuilara]
AxelPolleres: we expect that implementations support for functions, strongly support for funcitons to be accessible by keywords
15:33:43 [kasei]
AxelPolleres, how is this different from the xsd constructor functions we already require but don't have keywords?
15:33:52 [LeeF]
q?
15:33:53 [AndyS]
Some possible names: http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html
15:33:53 [cbuilara]
LeeF: SPARQL users do not realize of all the functions they have available
15:35:03 [cbuilara]
LeeF: Axel is in favor, Kasei has objections
15:35:06 [AndyS]
Yes - I'd put in names as per emaikl (not DateTime ones 'cos I can't think of names)
15:35:13 [LeeF]
http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html
15:35:22 [pgearon_]
+q
15:35:28 [LeeF]
ack pgearon
15:35:43 [cbuilara]
pgearon: rather not include a particular prefix
15:35:57 [cbuilara]
pgearon: just say date, day
15:36:16 [cbuilara]
pgearon: it should be common for all queries
15:36:31 [cbuilara]
LeeF: not in favor
15:36:36 [kasei]
pgearon, like virtuoso's bif: prefix?
15:36:51 [AndyS]
+1 to LeeF - principle is URIs - else keywords
15:37:31 [AxelPolleres]
implicit prefixes need some more things to define... although I am not strongly opposing that, that would cost more time to agree than we have, I am afraid... keywords are the shorter route, it seems
15:37:58 [cbuilara]
LeeF: any other proposal?
15:38:11 [kasei]
I'd support keywords for everything for simplicity as long as the keywords remain \w+ (with no dashes, underscores) including dateTime functions.
15:38:14 [chimezie]
Zakim, unmute me
15:38:14 [Zakim]
Chimezie_Ogbuji should no longer be muted
15:38:21 [AxelPolleres]
the keywords would just be the function names with fn: stripped off.
15:38:36 [cbuilara]
chimezie: any criteria on whihc functions have keywords?
15:38:41 [AndyS]
Axel - see email - not quite.
15:38:55 [cbuilara]
LeeF: multiword ones
15:39:25 [AxelPolleres]
CamelCase?
15:39:41 [kasei]
possible dateTime functions without multiwords: YEAR(), MONTH(), DAY(), HOURS(), MINUTES(), SECONDS(), TIMEZONE()
15:40:08 [AxelPolleres]
+1 to kasei's, sounds fine to me
15:40:13 [LeeF]
straw poll: (1) keywords as per AndyS's mail (excludes date time functions) (2) keywords for all (some keywords not yet specified anywhere)
15:40:26 [chimezie]
1
15:40:28 [cbuilara]
LeeF: strawpoll, two choices 1- keywords as per AndyS 2- keywords for all some not specified anywhere
15:40:31 [AxelPolleres]
2 ... kasei's proposal on IRC seems just fine
15:40:34 [sandro]
0 can't decide
15:40:42 [LeeF]
0
15:40:44 [cbuilara]
0
15:40:44 [AndyS]
0 - don't care
15:40:47 [MattPerry]
0
15:40:53 [AndyS]
2
15:41:09 [kasei]
0 - neutral after this discussion
15:41:11 [pgearon_]
2
15:41:15 [bglimm]
0
15:41:17 [OlivierCorby]
1
15:41:20 [AlexPassant]
0
15:41:53 [cbuilara]
OlivierCorby: maybe we align with xquery
15:42:15 [AxelPolleres]
In xquery, I can use the fn: functions without prefix, BTW.
15:42:25 [cbuilara]
OlivierCorby: problem: introducing more and more keyword, maybe to align to xquery
15:43:36 [cbuilara]
LeeF: slight preference to include keywords for all of this
15:43:43 [AndyS]
?? Xquery uses base URI resolution to make URIs of name without ":"
15:44:21 [LeeF]
PROPOSED: SPARQL 1.1 includes keywords for all functionsin function library, as per suggestions in http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html and using YEAR, MONTH, DAY, HOURS, MINUTES, SECONDS, TIMEZONE for the date time accessor functions
15:44:45 [LeeF]
PROPOSED: SPARQL 1.1 includes keywords for all functionsin function library, as per suggestions in http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html and using YEAR, MONTH, DAY, HOURS, MINUTES, SECONDS, TIMEZONE for the date time accessor functions and RAND
15:44:56 [pgearon_]
seconded
15:45:24 [LeeF]
RESOLVED: SPARQL 1.1 includes keywords for all functionsin function library, as per suggestions in http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html and using YEAR, MONTH, DAY, HOURS, MINUTES, SECONDS, TIMEZONE for the date time accessor functions and RAND
15:45:25 [cbuilara]
LeeF: anybody objects or abstain?
15:45:33 [NickH]
yay
15:46:03 [kasei]
no, but we've recently got this: http://www.w3.org/ns/sparql#
15:46:04 [LeeF]
http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html
15:46:06 [cbuilara]
LeeF: what URIs to use? SPARQL 1.0 has well defined URIs?
15:46:12 [cbuilara]
AndyS: no
15:46:19 [AxelPolleres]
Andy, does Xquery say something about default for the BASE URI to fn: (... at least more than one XQuery processr I played around with did that)?
15:46:56 [cbuilara]
AndyS: all SPARQL 1.0 in SPARQL 1.1 with their own namespace
15:47:34 [AndyS]
Axel - would need to check the details for XQ formal spec.
15:47:36 [LeeF]
zakim, who's talking?
15:47:46 [Zakim]
LeeF, listening for 10 seconds I heard sound from the following: LeeF (80%), Sandro (84%)
15:47:47 [chimezie]
Zakim, mute me
15:47:47 [NickH]
Lots of echo!
15:47:49 [Zakim]
Chimezie_Ogbuji should now be muted
15:47:58 [LeeF]
zakim, mute sandro please
15:47:58 [Zakim]
Sandro should now be muted
15:48:26 [LeeF]
http://www.w3.org/ns/sparql#
15:48:55 [cbuilara]
LeeF: use http://www.w3.org/ns/sparql# namespace for SPARQL 1.0 functions, any objection?
15:49:08 [AndyS]
q+ to ask about mapping to F&O (can wait)
15:49:31 [LeeF]
sandro: is that the same as service description?
15:49:38 [LeeF]
kasei: no, not the same - similar
15:49:38 [AxelPolleres]
andy, hmmm, in http://www.w3.org/TR/xquery-semantics/ I didn't find anything indicating that...
15:49:43 [LeeF]
sandro: should it be sparql-functions#?
15:49:58 [AndyS]
Axel - try the grammar/syntax.
15:50:24 [kasei]
service description namespace is: http://www.w3.org/ns/sparql-service-description
15:50:33 [LeeF]
sandro: or sparql-function# ?
15:50:46 [LeeF]
AndyS: need prefixes also for entailment regimes, ... what else?
15:51:01 [cbuilara]
LeeF: unclear when naming aggregates
15:51:03 [bglimm]
profiles in ent. regimes (OWL EL, RL, QL)
15:51:16 [bglimm]
that's used in the service descriptions
15:51:45 [cbuilara]
AndyS: put all service descriptions in one wikipage
15:51:51 [cbuilara]
AndyS: all prefixes
15:51:51 [LeeF]
AndyS: Can someone pull namespace issues together in one place?
15:51:59 [AxelPolleres]
we have yet another namespace for ent-regiems, BTW, right?
15:52:13 [cbuilara]
kasei: I can try to sumarize
15:52:37 [AxelPolleres]
items that need URIs and which namespace we currently use.
15:52:51 [LeeF]
ACTION: Greg to summarize what items we need URIs for and what URI prefixes we're currently using
15:52:51 [trackbot]
Created ACTION-338 - Summarize what items we need URIs for and what URI prefixes we're currently using [on Gregory Williams - due 2010-12-07].
15:53:25 [LeeF]
PROPOSED: All SPARQL 1.0 and SPARQL 1.1 functions have URIs in a SPARQL-owned namespace to be determined
15:54:03 [AndyS]
q-
15:54:17 [kasei]
+1 LeeF
15:54:32 [kasei]
q+
15:54:33 [cbuilara]
LeeF: any objection
15:54:37 [LeeF]
ack kasei
15:54:49 [AndyS]
LeeF: Simplist to have all functions in one space
15:55:26 [AxelPolleres]
+1
15:55:26 [AndyS]
FYI: Can call aggs by URI currently
15:55:30 [LeeF]
RESOLVED: All SPARQL 1.0 and SPARQL 1.1 functions have URIs in a SPARQL-owned namespace to be determined
15:56:22 [cbuilara]
LeeF: remaining issues, simple literal, access string stuff
15:56:42 [cbuilara]
LeeF: remaining: string concatenation
15:56:50 [cbuilara]
LeeF: latests proposals from AndyS
15:56:54 [LeeF]
http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0283.html -- has a proposal for how to handle simple literals, xsd:strings, and language tags in concat
15:57:00 [AndyS]
If we have CONCAT, I'm happy to drop overloading "+"
15:57:13 [LeeF]
http://lists.w3.org/Archives/Public/public-rdf-dawg/2010OctDec/0302.html - proposes not to require overloading + for concatenation
15:57:24 [AxelPolleres]
"CONCAT("a"^^xsd:string, "b") -> "ab"^^xsd:string (?? choice point)" ...
15:57:45 [AxelPolleres]
the coin flip is basically whetherr to default to return type plain lit, or string, isn't it?
15:57:50 [cbuilara]
LeeF: people to catch up in the mail threads, comments please
15:58:48 [chimezie]
bye
15:58:50 [Zakim]
-Chimezie_Ogbuji
15:58:51 [Zakim]
-LeeF
15:58:51 [Zakim]
-MattPerry
15:58:51 [bglimm]
bye
15:58:52 [Zakim]
-AxelPolleres
15:58:53 [Zakim]
-AlexPassant
15:58:56 [Zakim]
-bglimm
15:58:57 [MattPerry]
\quit
15:58:57 [Zakim]
-Sandro
15:58:57 [Zakim]
-pgearon
15:58:58 [Zakim]
-OlivierCorby
15:58:59 [Zakim]
-AndyS
15:59:01 [Zakim]
-NicoM
15:59:05 [Zakim]
-cbuilara
15:59:07 [Zakim]
-NickH
15:59:09 [Zakim]
-kasei
15:59:11 [Zakim]
SW_(SPARQL)10:00AM has ended
15:59:13 [Zakim]
Attendees were cbuilara, +1.617.553.aaaa, LeeF, AndyS, kasei, NickH, +49.911.973.4.aabb, NicoM, OlivierCorby, MattPerry, AxelPolleres, +1.540.841.aadd, Sandro, bglimm, pgearon,
15:59:16 [Zakim]
... AlexPassant, Chimezie_Ogbuji
16:00:22 [OlivierCorby]
OlivierCorby has left #sparql
16:12:15 [cbuilara]
cbuilara has joined #sparql
16:16:24 [AxelPolleres_]
AxelPolleres_ has joined #sparql
16:59:17 [AndyS]
SUBSTRING or SUBSTR? (we have STR)
17:09:58 [AndyS]
LENGTH or STRLEN ? Put STR* on all string functions (except SUBSTR...)
17:14:57 [AndyS]
STRLEN seems to avoid confusion with LENGTH of list, seq, path, etc
17:15:01 [AndyS]
kasei
17:20:24 [pgearon_]
pgearon_ has joined #sparql
17:21:02 [AndyS]
CONTAINS might also be confusing but alts aren't nice. STRCONTAINS is a bit long.
17:23:22 [AndyS]
CEIL vs CEILING?
17:25:45 [kasei]
AndyS?
17:26:08 [kasei]
agree STRCONTAINS feels long.
17:26:17 [kasei]
I'm happy with CEIL.
17:26:40 [kasei]
Could go either way on SUBSTR(ING) and LENGTH/STRLEN.
17:28:14 [AndyS]
CEIL then
17:28:20 [AndyS]
LENGTH is now looking scary (= trouble causing :-) given lists, seq, bag
17:28:46 [AndyS]
CONTAINS is sort-of OK. More stringish than listish (member, index)
17:30:57 [kasei]
if you're worried about it, let's do STRLEN then.
17:31:55 [kasei]
I'm not a huge fan, but I agree it might be confused with list/seq ops.
17:33:35 [AndyS]
Maybe its because I was thinking about the path comment I need to reply to.
17:36:32 [kasei]
not sure if you've got any of these implemented in arq, but was hoping you might take a look at the function tests I committed at some point.
17:37:28 [kasei]
as for string ops, I think I'm leaning towards xsd:string if all args are xsd:string, and similarly for language tags. dropping back to simple literals otherwise (no errors as long as the args are literals).
17:52:50 [AndyS]
I have just found I do have most of them -- didn't realise I'd added fn:XYZ versions before. My versions are all simple literal returning but will add xsd:string, lang versions to follow eventual WG decision.
17:53:24 [AndyS]
ARQ also has xsd:date and xsd:g*.
17:55:07 [AndyS]
Tricky case: starts("foo"@en, "f") :: I say arg must be pure characters -> simple + xsd:string only. Avoid ss and ß
17:58:51 [Zakim]
Zakim has left #sparql
17:58:53 [kasei]
both args must be just chars?
17:59:00 [kasei]
that is a good (tricky) case
17:59:05 [kasei]
any notion of why fn:year-from-dateTime doesn't return a xsd:gYear?
17:59:31 [kasei]
brb
18:06:13 [cbuilara]
cbuilara has left #sparql
18:15:37 [AndyS]
Because then you need "get the year as a number" from xsd:gyear? Symmetry with other suggests number.
18:16:19 [AndyS]
Opps - dajobe not going to be keen on todays decisions :-)
18:16:31 [kasei]
yeah, I know
18:16:55 [kasei]
when I write it up for the namespace(s) page, should I list LENGTH or STRLEN?
18:20:05 [AndyS]
STRLEN
18:30:22 [kasei]
http://www.w3.org/2009/sparql/wiki/SPARQL_Namespaces
18:30:44 [kasei]
yikes. written out like that, it really does look like a lot of new keywords.
18:39:41 [kasei]
that's odd
18:39:52 [kasei]
isBlank is missing from the SPARQL 1.0 keywords table.
18:41:08 [kasei]
ah. i see it's been added ot the 1.1 draft.
18:41:48 [AndyS]
Was going to remove some duplication - there's more than one list of keywords.
18:42:32 [kasei]
oh, really?
18:48:02 [kasei]
i wonder how much use fn:timezone-from-dateTime gets.
18:49:17 [kasei]
seems like the odd one out for inclusion in sparql. easy enough to implement, but a rather odd return value unless you natively support durations.
19:11:31 [AndyS]
Maybe a different return for SPARQL? String (to cover Z)
19:14:14 [pgearon_]
pgearon_ has joined #sparql
19:15:57 [AndyS]
Anyone know what's the state of python, ruby, (insert lang here) for SPARQL 1.1?
19:26:47 [kasei]
only that redland has partial support and there are bindings for most of those languages...
19:26:59 [kasei]
and there's excellent support in perl :)