Semantic Web Interest Group IRC Chat Logs for 2005-03-08

00:12:00 <demaledetti> how can a SPARQL query retrieve all elements from an arbitrarily long rdfs:Collection?

00:14:49 <crschmidt> rdf:Collection

00:15:31 <demaledetti> oops, yeah thanks

00:15:33 <crschmidt> and, afaik, it can't. And I've asked SPARQL people about it a couple times with no response yet. You can look at all the triples pointed to by an rdf:first, which are all items of the Collection if there's only one in the model

00:17:27 <demaledetti> thought so...yet very surprising.. this doesn't make it very easy..

00:19:21 <sbp> <DAWG> lalala we are not listening

00:20:16 <demaledetti> :-) maybe a "in" operator a la python?

00:20:19 <sbp> it is on the issues list through

00:20:23 <sbp> er, though

00:20:32 <sbp> member would be a better name

00:20:35 <crschmidt> Yeah, but no discussion of it that I can find...

00:28:07 <demaledetti> I would be happy to say "ok who needs Collections anyway" if they weren't used as the range of so many useful OWL properties...

00:37:01 <sbp> and in obscure syndication formats

00:40:21 <demaledetti> yeah, hey this is starting to get funny, seems that you guys mean that it won't ever get solved in SPARQL... or at least that you are losing hope..

00:41:00 <sbp> we're just cynical

00:41:10 <sbp> it *should* get solved in SPARQL

00:41:14 <sbp> but I won't be surprised if it isn't

00:41:21 <demaledetti> that's what I meant

00:41:46 <sbp> there are always custom querying methods

00:42:20 <sbp> every half-decent RDF API has a low query interface, and it's not too difficult to build something SPARQL sized on top of even a rudimentary triple matcher

00:42:41 <sbp> making it efficient is the natcher

00:43:58 <sbp> with collection, you can't be all that efficient unless the internal representation specifically optimises collections since it's just a huge first/rest chain that you have to keep plumbing the depths of until you get to the required member

00:44:51 <demaledetti> so SPARQL is really trying to be like SQL.. a plethora of dialects, implemented with many different nuances in implementations mainly due to optimization constraints... amazing

00:44:57 <sbp> so as long as the API a) optimises internally and b) provides an interface to query using that optimisation you are, in the words of Morbus, Golden Grahams. and even if it doesn't you can usually get something going that is more than adequate on sub-megatriple stores

00:45:15 <sbp> well notice the QL at the end and continuing SQL-retro syntax

00:45:38 <sbp> some people just won't accept that disco is dead

00:45:49 <TML> Disco?

00:45:51 <sbp> <DiscoStu> if this trend continues... eeeehy!

00:45:59 <sbp> yes, that other wonderful product of the 70s

00:46:05 <TML> Ah. :)

00:46:08 <demaledetti> I hoped that meant "hey this is disco after learning the lesson"

00:48:11 <demaledetti> golden grahams? food?

00:48:36 <sbp> some American thing; I don't pretend to understand it

00:48:57 <crschmidt> It's not even American slang, it's Morbus slang.

00:49:13 <crschmidt> "You're golden" is american slang for "you're all set - no problems"

00:49:17 <sbp> golden is popularly used as an adjective for good... right

00:49:21 <crschmidt> Morbus just has to throw the grahms on the end, I assume.

00:49:51 <sbp> I believe golden is used on this side of the pond too

00:50:07 <crschmidt> ah

00:50:16 <crschmidt> wouldn't know. You rightpondians scare me

00:50:37 <demaledetti> given this is the state of the art you just gotta invent something to cheer you up I guess, thanks for the explanation :-)

00:51:03 <sbp> dextropondians

00:51:17 <crschmidt> and we're what? Sinisterpondians?

00:51:38 <sbp> you betcha

00:53:38 <crschmidt> even with SPARQL though, it's not that difficult to create a complete collection. As sbp mentioned, you just drill until you see an rdf:nil or whatever, so as an application-level device, it's pretty simple

00:54:11 <crschmidt> But thus far, there's been no real discussion of it for inclusion in SPARQL as far as I'm aware. (And I tend to notice these things, simply because I have a use for them ;))

00:54:58 <sbp> N3QL would have it by extension

00:56:07 <demaledetti> yeah call me lazy or wishing for a free lunch but I thought that inclusion in a list could've been easier for the ultimate W3C tool for querying knowledge bases, oh well, I guess I will adapt myself once the disappointment decreases...

00:56:44 <sbp> well wait until it's finished; and make a noise about it!

00:57:02 <demaledetti> I am trying to ;-)

00:57:03 <sbp> if you want the feature, ask for it. Last Call isn't even in sight yet, so it's not like there's no time

00:59:28 <demaledetti> hey ericP, I get back all the nice things I wrote about SPARQL earlier ;-P

01:08:22 <bjoern_> "Simply Put, Another Ratified Query Language"?

01:18:52 <Mutiny> less query languages, more data to query!

01:21:19 <demaledetti> the first is a precondition for the latter, the more fragmented, prematurely optimized and bound to specific APIs the query languages, the less queriable data

01:23:10 <crschmidt> ... the query language is completely unrelated to the data

01:23:15 <crschmidt> plus, there's plenty of data out there, trust me

01:23:16 <crschmidt> ^status

01:23:17 <julie> I currently hold 2124649 triples, 75 namespaces, and 60 stored commands. I have been running for 25.055 hours, and have used 0.095 minutes of CPU time. Stats on current machine (peanut.crschmidt.net): 20:30:46 up 1 day, 3:26, 6 users, load average: 0.22, 1.23, 1.70

01:23:53 <Mutiny> it's a pretty straight forward chicken and egg problem :/

01:24:15 * crschmidt doesn't understand it that way, but could be missing something

01:24:18 * demaledetti , instead of publishing data tonight, is redesigning it to better suit the query language

01:25:42 * demaledetti wishes the query language was unrelated enough to the data as to not make it impossible to query it

01:40:05 <chillywilly> lalala

01:40:19 <chillywilly> herro

01:41:29 <chillywilly> is their a more appropriate channel for discussing al thing REST?

01:41:34 <chillywilly> all*

01:41:41 <chillywilly> bleh...

01:42:44 <dviner> if you ask your question, i expect someone will either answer or offer a better channel

03:17:49 <bparsia> bparsia is now known as bijan

03:39:50 <ericP> demaledetti, sbp, crschmidt, anybody else who gives a damn, we gave up on collections for this iteration 'cause of closure issues.

03:39:59 * ericP hunts down a relevent thread...

03:43:07 <ericP> querying over collection; woes --> http://lists.w3.org/Archives/Public/public-rdf-dawg/2005JanMar/thread.html#83

03:45:10 * bijan cheers!

06:31:27 <kasei> anyone around who knows the relationship schema?

06:32:54 * CloCkWeRX tentatively proffers a hand

06:33:22 <kasei> i'm reading the spec, and worried that i've been using some of the predicates backwards

06:33:36 <kasei> parentOf, for example is "A person who has given birth to or nurtured and raised this person."

06:34:13 <kasei> that seems backwards to me, and from what I can tell, is used wrong in several results from julie's dataset

06:34:38 <kasei> although I'm to blame for a lot of what julie is giving back

06:35:43 <kasei> any thoughts?

07:20:01 <deltab> kasei: is there any comparison between birth dates that you can use to disambiguate?

07:22:25 <kasei> not that i can see

07:22:43 <kasei> at least not in the spec.

07:23:37 <kasei> but in the data from julie, it looks like the opposite use of the spec text. (ie, it's being used so that ?p rel:parentOf ?q ==> ?p olderThan ?q)

08:54:34 <[GNU]> moin

09:34:19 <PhilT> Has sh1mmer been on line today?

09:44:22 <eaon> no

10:11:25 <[GNU]> [GNU] is now known as gnu_jabber

11:06:42 <bengtf> where can I find a owl dtd or relax ng ?

11:09:01 <kota> it shouldn't be dtd nor relaxng...

11:10:02 <bengtf> anything then ....

11:10:23 <bengtf> need for emacs nXML mode

11:15:15 <kota> oh there's owl-mode http://owl-emacs.projects.semwebcentral.org/ . this might help...

11:16:06 <bengtf> thanks

12:22:13 <reto> hello?

12:22:57 <eaon> hey reto

13:21:51 <chillywilly> I have a prodct (metal) that is categorized as follows: material->form->grade, how can I make a RESTful catalog? I've been thinking of doing something like http://www.oursite.com/carbon_steel_cold_rolled/Flat/1018 (material/form/grade) or do you guys think a query string would be better (sory not that big a fan of using the query string ;P)

13:23:20 <chillywilly> this is also further subdividd into sizes like: 1/8 x 1/4, 1/8 x 5/16, etc. I am sure this is a common thing as lots of ecommerce apps categorize their products

13:23:27 <chillywilly> subdivided too

13:24:12 <chillywilly> or well a lot of *companies* categorize their products I should say

13:25:52 <bijan> Looks fine to me

13:26:05 <bijan> No need for a query string

14:50:28 <bengtf__> bengtf__ is now known as bengtf

14:55:18 <DanC_lap> .t

14:55:20 <phenny> Tue, 08 Mar 2005 14:55:19 GMT

14:55:40 <DanC_lap> hmm... this hotel net connection is so flakey... I can't seem to make an ssh connection to the office...

14:56:36 <LotR> that's just life telling you to take the day off :)

15:22:54 <demaledetti> ericP: thanks for the pointer, I will try and understand that

15:23:32 * demaledetti promises not to be noisy about the collection access issue again until he has a better understanding of it...

17:03:39 <rreck> I strongly recommend avoiding work with McGraw Hill

17:06:25 <libby> what brought that on rreck?

17:06:33 <rreck> frustration

17:06:54 <rreck> 1099 me on my expenses then telling me to file schedule C to remove their mistakes

17:07:15 <rreck> unapplogetically i might add

17:08:51 <rreck> i could provide a litany of grievances but im sure noone is interested

17:20:14 <DanC_lap> hmm... lunchtime here at the IETF... I wonder about setting up wordpress for www.webtechs.com

17:21:23 <mortenf> please go for new 1.5 and tell us of your experience :)

17:23:55 <DanC_lap> DanC_lap is now known as DanC_MSP

17:25:06 <crschmidt> 1.5 is nice, in my experience

17:25:13 <crschmidt> Set it up for Jess and myself, no problems

17:34:25 <nmg_> nmg_ is now known as nmg

19:03:33 <DanC_MSP> BLURB: SIP Session 2

19:03:58 <DanC_MSP> heeeere, dc_swig bot... heeere bottie bottie...

19:04:47 <DanC_MSP> hmm... I'm here at IETF 62 in the SIP WG.

19:06:39 <DanC_MSP> rosenberg presents some slides that refer to http://www.jdrosen.net/papers/draft-rosenberg-sip-target-dialog-00.txt

19:06:57 <DanC_MSP> "JDR write up and posts URL for target dialog draft"

19:07:05 * DanC_MSP is struggling to figure out which end is up

19:07:11 <DanC_MSP> logger, pointer?

19:07:12 <DanC_MSP> See http://www.ilrt.bris.ac.uk/discovery/chatlogs/swig/2005-03-08#T19-07-11

19:07:27 <DanC_MSP> ok, my notes won't be forgotten, anyway

19:07:51 <mortenf> dajobe was here 15 minutes ago, can't be far...

19:08:09 <dajobe> hello?

19:08:17 <mortenf> no bot...

19:08:18 <dajobe> ok

19:08:26 * DanC_MSP finds SIP meeting agenda http://www.ietf.org/ietf/05mar/sip.txt

19:09:05 * DanC_MSP finds SIP WG charter http://www.ietf.org/html.charters/sip-charter.html

19:09:31 <DanC_MSP> ugh.. speakers don't identify themselves at the mic, despite request from chair

19:10:01 <DanC_MSP> heaven forbid the internet draft should point to the WG charter, nor that the WG charter should point to the meeting agenda.

19:11:22 <DanC_MSP> lots of discussion of "target dialog"...

19:11:39 <DanC_MSP> chair puts a question by humm... proposal is displayed

19:11:44 <DanC_MSP> but I think it's in ppt somewhere.

19:12:01 * DanC_MSP wonders what visually disabled folks do at IETF meetings

19:12:20 <DanC_MSP> I think there's an XMPP channel somewhere... I wonder if it's used...

19:12:34 <dajobe> dc_swig: view 1

19:12:41 <DanC_MSP> "Outbound Connections" [crud; missed the name]

19:12:50 <dajobe>http://www.ietf.org/

19:12:50 <dc_swig> A: http://www.ietf.org/ from dajobe

19:12:53 <dajobe> A:|IETF

19:12:53 <dc_swig> Titled item A.

19:12:58 <DanC_MSP> ah... Cullen Jennings

19:13:18 <DanC_MSP> BLURB: SIP Session 2

19:13:18 <dc_swig> B: SIP Session 2 from DanC_MSP

19:13:39 <DanC_MSP> B:|SIP Session 2 at 62nd IETF in Minneapolis, MN

19:13:39 <dc_swig> Titled item B.

19:13:53 <DanC_MSP> B:[http://www.ietf.org/ietf/05mar/sip.txt|meeting agenda]

19:13:54 <dc_swig> Added comment B1.

19:14:04 <DanC_MSP> B:[http://www.ietf.org/html.charters/sip-charter.html|wg charter]

19:14:04 <dc_swig> Added comment B2.

19:15:12 <DanC_MSP> B:[http://www.jdrosen.net/papers/draft-rosenberg-sip-target-dialog-00.txt|draft-rosenberg-sip-target-dialog-00] one of the drafts on the agenda (missed IETF -00 deadline (ala w3c publication moratorium deadline))

19:15:12 <dc_swig> Added comment B3.

19:15:52 <DanC_MSP> B1:[http://www.ietf.org/ietf/05mar/sip.txt|SIP meeting agenda]

19:15:52 <dc_swig> Replaced comment B1.

19:16:03 <DanC_MSP> B2:[http://www.ietf.org/html.charters/sip-charter.html|SIP wg charter]

19:16:03 <dc_swig> Replaced comment B2.

19:17:14 <DanC_MSP> B: heaven forbid the internet draft should point to the WG charter, nor that the WG charter should point to the meeting agenda, nor that the agenda should link to the docs

19:17:15 <dc_swig> Added comment B4.

19:18:08 <DanC_MSP> B:let alone accessible presentation materials. The chair put a question ala "any objections to what's displayed [in .ppt, not available by http]? no? ok, humm of support? ok, scribe, record that."

19:18:08 <dc_swig> Added comment B5.

19:18:42 <DanC_MSP> B:"URI tag for STUN support (like sigcomp)" ugh.. struggling to de-code jargon

19:18:42 <dc_swig> Added comment B6.

19:19:22 <dajobe> that's one of the sip across NAT things, I know skype uses 2 of them and others

19:19:49 <DanC_MSP> B:at least 100 people in the room. folks at the mic must be reading the drafts in advance and such. I wonder about the rest

19:19:49 <dc_swig> Added comment B7.

19:24:17 <DanC_MSP> ah... now the chair is dealing with dissent...

19:25:33 <DanC_MSP> guy at the mic asks how many have read it... "not too many... this is the most important SIP extension on the agenda... everybody's doing these horrible registration hacks... this is why it's so difficult to get a new client talking to your SIP gateway [?]"

19:25:55 <DanC_MSP> er... discussion moves on. I wonder who has the ball there.

19:26:49 <DanC_MSP> I wonder... how much is SIP different from XMPP?

19:27:27 <DanC_MSP> another topic...

19:27:46 <DanC_MSP> ?: "that's something people are going to get wrong, unless there's some very, very clear text about that"

19:28:02 <DanC_MSP> hmm... expecting developers to read specs... how quaint.

19:28:11 <DanC_MSP> no discussion of test cases. sigh.

19:29:35 <DanC_MSP> discussion of pub/sub of certificates...

19:30:04 <DanC_MSP> "Certificate Management"

19:30:58 <DanC_MSP> B:hmm... where is SIP w.r.t its schedule? sigh... charter thinks "Mar 04" is in the future.

19:30:58 <dc_swig> Added comment B8.

19:38:47 <[GNU]> hello all.

19:41:18 <[GNU]> anybody knows it the planetrdf.com aggregator code is available?!

19:41:44 <DanC_MSP> B:wild... this SIP WG is doing all kinds of stuff... [Hypertext Transfer Protocol (HTTP) Digest Authentication Using Authentication and Key Agreement (AKA) (RFC 3310)|http://www.ietf.org/rfc/rfc3310.txt] is one of their RFCs

19:41:44 <dc_swig> Added comment B9.

19:41:52 <deltab> [GNU]: assuming it's Planet: http://planetplanet.org/

19:41:52 <dajobe> [GNU]: Yes I know. Yes it is.

19:41:58 <dajobe> deltab: no

19:43:10 <[GNU]> dajobe: oh, planetplanet would run on my machine :(

19:43:14 <dajobe> [GNU]: see http://www.hackdiary.com/archives/000036.html

19:44:02 <[GNU]> ahm, thats about the chumpbot?!

19:44:06 <DanC_MSP> er... did the SIP WG charter/CFP notify the community that they would be working on HTTP auth? nope... the only mention of HTTP in the charter is "SIP is a text-based protocol, similar to HTTP and

19:44:07 <DanC_MSP> SMTP..."

19:44:17 <DanC_MSP> logger, chump B:

19:44:17 <DanC_MSP> B:See [http://www.ilrt.bris.ac.uk/discovery/chatlogs/swig/2005-03-08#T19-44-17|discussion]

19:44:17 <dc_swig> Added comment B10.

19:44:18 <dajobe> [GNU]: no, read the title

19:44:34 <deltab> dajobe: no it's not Planet?

19:44:44 <dajobe> for the third time. no.

19:44:45 <[GNU]> i think so to

19:44:51 <[GNU]> thats not the planetrdf.com stuff

19:45:39 <[GNU]> maybe chumpologica is even better... :)

19:45:59 <dajobe> for the fourth time. chumpologica is the planetrdf stuff

19:46:21 <mortenf> read the title, skip first line to avoid confusion...

19:48:49 <dajobe> right, I changed the sidebar to say that

19:52:15 <[GNU]> hmm, thats a better start than planetplanet

19:52:21 <[GNU]> thanks for bearing with me

19:56:34 <deltab> dajobe: sorry about that

19:56:51 <DanC_MSP> ugh... I'm having trouble with the XMLRPC interface to my gizmo data

19:56:57 <dajobe> deltab: s'okay, it wasn't entirely clear

19:57:20 <crschmidt> DanC_MSP: the sidekick?

20:00:52 <DanC_MSP> right, sidekick

20:01:10 <DanC_MSP> GPRS isn't working here in Minneapolis

20:01:36 <DanC_MSP> and I've noticed a lot more "web browser service has timed out" problems lately.

20:02:22 <DanC_MSP> I'm starting to wonder if the dngr.com services are sufficiently scalable

20:03:18 <DanC_MSP> (lots of talk about MIME type granularity here in SIP... and now discussion of an XML namespace... where are we?)

20:03:53 <DanC_MSP> "Proposal: [darn! slide moved too fast" they're mucking with the Accept header.

20:04:57 <crschmidt> DanC_MSP: Danger's network has been having problems over the past two days, according to T-mobile customer support, and they're "working on it"

20:05:25 <crschmidt> Not sure how much that relates to your specific issue: I wouldn't expect a GPRS/gizmo level problem to carry over into XMLRPC, but have had a couple friends with sidekicks mention it lately

20:16:15 <DanC_MSP> eek... urn:ietf:params:xml:ns:pidf:geopriv10;

20:16:19 <DanC_MSP> "Comments?"

20:16:31 <DanC_MSP> "Civil Location Format"

20:16:45 <DanC_MSP> ... geopriv10:civilLoc

20:19:15 <danbri> oh, interesting. Geopriv was looking worth investigating last I heard a talk about it... but I've not made time yet.

20:24:15 <DanC_MSP> B:long discussion of negotiating extensions... reminds me of "untested hooks considered harmful" rant

20:24:15 <dc_swig> Added comment B11.

20:26:31 <DanC_MSP> B:[http://www.xmpp.org/ietf-logs/sip@ietf.xmpp.org/2005-03-08.html|sip@ietf.xmpp.org log] shows some activity, so I tried to connect to the channel, but failed

20:26:31 <dc_swig> Added comment B12.

20:28:27 <DanC_MSP> discussion of SIP proxies... seems to be a late addition to the agenda...

20:30:37 <DanC_MSP> B:RFC3261 is referenced often

20:30:38 <dc_swig> Added comment B13.

20:31:02 <DanC_MSP> B:SIP: Session Initiation Protocol (RFC 3261)

20:31:02 <dc_swig> Added comment B14.

20:31:56 <DanC_MSP> B:269 pages

20:31:56 <dc_swig> Added comment B15.

21:09:45 <DanC_MSP> B:adjourned.

21:09:46 <dc_swig> Added comment B16.

21:22:20 <DanC_MSP> ah... I'm not alone with my sidekick problems. http://www.hiptop.com/forums/viewtopic.php?t=26204

21:31:08 <michelp> michelp is now known as michelp|town

21:46:47 <jsled> DanC_MSP: http://engadget.com/entry/1234000337034991/

21:51:32 <DanC_MSP> ugh

21:52:12 * DanC_MSP googles for "series 60 device" and the #1 hit is about python support. maybe the gods are speaking to me

21:52:15 <DanC_MSP> . http://www.forum.nokia.com/main/0,,034-821,00.html

21:53:09 <DanC_MSP> hmm... no thumb-keyboard devices...

21:55:59 <crschmidt> .g symbian python

21:56:03 <phenny> symbian python: http://crschmidt.net/blog/categories/mobile-platform/symbian-python/

22:20:32 * DanC_MSP fails to avoid spending lots of time exploring symbian-python

22:20:58 <crschmidt> heh

22:23:15 <crschmidt> did you find the postneo wiki?

22:27:00 <DanC_MSP> no

22:27:12 <DanC_MSP> ooh... now that felt constructive... I just reported an outage via http://www.signalstrength.org/index.html

22:28:33 <crschmidt> .g postneo wiki series 60

22:28:41 <phenny> postneo wiki series 60: http://www.postneo.com/postwiki/moin.cgi/PythonForSeries60

22:28:43 <crschmidt> probably the best source on the Series 60 on Python stuff that's out there

22:34:07 <DanC_MSP> stop it stop it STOP IT! no more tasty links!!!

22:34:15 <DanC_MSP> ;-)

22:34:47 * DanC_MSP is playing window-whack-a-mole and losing badly

22:34:52 <jsled> heh.

22:35:18 <DanC_MSP> I have an article I'm sorta trying to write... trying to clear my desktop/mind first

22:35:29 <DanC_MSP> of course, chatting in IRC is an excellent way to do this. not.

22:46:46 <DanC_MSP> in the session I'm in now, the chair explicitly asked for a jabber scribe for remote participants. there were no volunteers, so it was dropped

22:47:45 <DanC_MSP> I think I'm in MAGMA (salon E)

22:56:44 * DanC_MSP switches to opes...

22:56:45 <DanC_MSP>http://www.ietf.org/ietf/05mar/opes.txt

22:56:45 <dc_swig> C: http://www.ietf.org/ietf/05mar/opes.txt from DanC_MSP

22:57:06 <DanC_MSP> C:|Open Pluggable Edge Services WG (opes) Tuesday, March 8 at 1645-1745

22:57:06 <dc_swig> Titled item C.

22:57:35 <DanC_MSP> C:[http://www.ietf.org/html.charters/opes-charter.html|Open Pluggable Edge Services (opes) charter] Last Modified: 2005-01-07

22:57:35 <dc_swig> Added comment C1.

22:57:56 <DanC_MSP> C:presentation suggests HTTP part is done; moving to SMTP; looking at use cases

22:57:56 <dc_swig> Added comment C2.

22:58:51 <DanC_MSP> C:wg charter says "

22:58:51 <DanC_MSP> The OPES Working Group has previously developed an architectural framework to authorize, invoke, and trace such application-level services for HTTP."

22:58:52 <dc_swig> Added comment C3.

22:59:07 <DanC_MSP> C3:charter says "The OPES Working Group has previously developed an architectural framework to authorize, invoke, and trace such application-level services for HTTP."

22:59:07 <dc_swig> Replaced comment C3.

23:10:55 <eaon> hey sh1mmer