Experimental IRC log swig-2007-09-14

Available formats: content-negotiated html turtle (see SIOC for the vocabulary)

Back to channel and daily index: content-negotiated html turtle

These logs are provided as an experiment in indexing discussions using IRCHub.py, Irc2RDF.hs, and SIOC.

00:26:20<drobilla>hm.. can you express /anything/ in RDF?
00:27:29<kpreid>two kinds of "not really":
00:27:48<kpreid>1. expressing, say, a megapixel image in RDF would be a bad idea
00:28:08<drobilla>I mean, theoretically speaking
00:28:10<kpreid>2. arguably you can't express something if there is nothing that understands the vocabulary you're using.
00:28:16<kpreid>but that lends itself to extension
00:28:30<drobilla>if there's no known counter-example, probably yes
00:28:49<kpreid>I think it's much like turing-completeness in programming languages
00:29:08<drobilla>yeah. that I can prove, but I'm no logician. no idea how to go about proving this :)
00:29:30<kpreid>insofar as "ability to express" is meaningful, RDF can express anything; but some things are more or less practical
00:31:53<drobilla>right. as you said, megapixel images in rdf is probably a very stupid thing to do. but you /can/
00:32:17<drobilla>I guess this is equivalent to asking "is all knowledge representable as relationships between entities"
00:32:42<drobilla>which might get philosophical quickly :)
00:33:39<kpreid>indeed
00:34:32<kpreid>what I'd watch out for is those situations where, in order to produce that form, the entities being referred to are not the entities of interest, but rather some meta-layer necessary to represent them in RDF
00:35:17<kpreid>for example, you can't write a >2-ary relation directly in RDF; you have to introduce a bnode to stand for the relationship and link each of the objects to it
00:35:36<kpreid>that's relatively harmless, but that's the sort of symptom I'm talking about
00:43:13<drobilla>yeah. far from the ideal model for everything
01:34:19<Arnia>kpreid: Interesting, that's entirely how the maths does it
01:37:59<c>Arnia: whats CBD?
01:38:17<c>"limited to the CBD of the item"
01:38:26<c>class by domain?
01:38:29<c>ACTION clueles
01:42:07<kpreid>c: per what Arnia's talked about before,
01:42:10<kpreid>.g "concise bounded description"
01:42:13<phenny>kpreid: http://www.w3.org/Submission/CBD/
01:42:32<kpreid>Arnia: which, the bnode thing?
01:42:59<c>thanks, the acronym eluded me :)
01:43:20<Arnia>kpreid: yeah, it is essentially currying... which is a requirement for turing-completeness
01:43:40<Arnia>c: Sorry, I tend to use acronyms when I can't keep up with the typing
01:44:29<Arnia>kpreid: In fact, RDF is sort of like a fully curried topic map
01:49:57<kpreid>interesting
01:52:14<Arnia>Yes, and people say my category theory excursions are esoteric ;)
02:51:08<CaptSolo>hi all
02:52:37<Arnia>Hey CaptSolo
02:54:56<CaptSolo>hi Arnia
02:54:59<CaptSolo>what's new?
02:57:02<Arnia>writing an ontology documentation tool as a test of my RDF library
02:59:09<Arnia>Yourself?
03:01:09<CaptSolo>Arnia: what's the documentation tool doing?
03:01:50<CaptSolo>ACTION wonders how is that similar (or not) to what SpecGen4 is doing (it's used to generate SIOC documentation)
03:03:01<CaptSolo>Arnia: i am currently revising a paper written earlier
03:03:32<CaptSolo>should be sleeping now perhaps, but the day was not very productive so am catching up at night-time
03:05:04<Arnia>CaptSolo: Well it will take an RDF/XML file and produce an XHTML file documenting the ontology with embedded RDFa
03:05:21<Arnia>CaptSolo: at least that's the plan. There is a secret feature planned too
03:05:32<CaptSolo>aha! sounds nice
03:05:53<CaptSolo>my tool does the first part - generating XHTML file.
03:05:57<CaptSolo>no embedded RDFA
03:06:25<CaptSolo>(it's demand driven, so when someone needs e.g. RDFa that's when it's added)
03:06:38<Arnia>Going to use jQuery to add a nice browsing interface too
03:07:05<CaptSolo>if you can give me a link to an ontology I could show what it generates. if that's of any help.
03:16:49<Arnia>CaptSolo: Just pick any I think... it would be interesting to compare outputs once I've finished
03:24:34<CaptSolo>Arnia: let me know when you're finished then. :)
03:25:25<CaptSolo>as for picking any - you can just look at SIOC spec
03:26:28<CaptSolo>ACTION goes into master procrasator mode by launching an episode of Dr.Who
03:26:36<CaptSolo>procrastinator
03:26:42<CaptSolo>that's better :)
03:27:41<Arnia>heh. Yay for procrastination ;)
03:28:03<timbl>Top ten reasons for procrastination:
03:28:03<timbl>1.)
03:29:16<CaptSolo>:D
03:29:28<CaptSolo>top 5, 10, ... list are popular these days
03:30:31<dajobe>#1 The Internet
03:31:15<timbl>ACTION gtg
03:55:13<Arnia>Wow, almost 5am already
04:08:34<CaptSolo>indeed :)
04:08:44<CaptSolo>we must be on the same tz
04:14:30<Arnia>Well, I'm in the UK and I presume you're in Ireland atm
04:15:15<CaptSolo>right. not that you'd tell that by asking time to my irc client :D
04:15:47<CaptSolo>isn't it a bit late / early now?
04:19:34<Arnia>I slept this evening... my body clock is always a little awry
04:19:38<Arnia>What about you? :)
04:25:04<CaptSolo>just had an unproductive day (should have stayed up last night but went to sleep instead hoping the morning will be ok)
04:25:30<CaptSolo>maybe it's change from summer to autumn that makes it that, dunno
04:25:52<CaptSolo>anyway, i am better at working in some far away tz :)
04:25:58<Arnia>Yeah... ach well.
04:26:08<Arnia>I'm just naturally nocturnal. Have been since I was a baby
04:26:14<CaptSolo>same here
04:26:51<Arnia>Without daily social contact in real life (such as when Durham is empty during the summer) I drift into a nocturnal pattern
04:27:05<CaptSolo>best time to work is at night. might not fit with others' schedules.
04:27:32<CaptSolo>social contact forces / brings you back to "normal"?
04:28:17<CaptSolo>heh, we're almost getting into philosophical questions (eg, what is normal) at half past 5 at night ;)
04:30:06<Arnia>CaptSolo: yes... part of the problem is how little I need to sleep a day. Five hours is my preferred sleep period
04:30:15<Arnia>CaptSolo: but I can do fine on four
04:33:51<CaptSolo>just 4 hours, continuously from day to day ?
04:34:08<dajobe>you could get a 2nd job
04:34:55<Arnia>CaptSolo: yeah... how do you think I'm able to have an active social life and do a PhD at the same time? ;)
04:34:58<CaptSolo>a friend once told me about an "uber-man" sleep schedule which consists of 0.5h naps throughtout the day and then one longer (2 hour?) period
04:35:16<Arnia>dajobe: be hell to figure out my tax though :)
04:35:22<CaptSolo>that might be quite rough though
04:36:13<CaptSolo>Arnia: let's have a drink to an active social life :)
04:37:53<Arnia>Yeah. I'm a 'work hard, play hard' sort of person. Can't stand to be cooped up for more than a few days.
04:38:24<Arnia>When I visited Galway, I made sure to have a look around and make sure there were plenty of pubs :)
04:42:11<CaptSolo>Galway's ok in terms of the number of pubs around here.
04:43:18<CaptSolo>semweb/foaf hackers had a chance to check them out during FOAF Galway (a long time ago)
04:43:43<Arnia>I did like the Quays
04:53:35<CaptSolo>ok, next time we should definitely meet up
05:13:42<Arnia>Indeed
07:23:34<EtnaRosso>morning all
08:21:06<rawles>morning - does anyone have any experience of editing a metaclass in Protege-OWL? I can't seem to work out how its done
09:45:04<EtnaRosso>morning
09:45:35<EtnaRosso>using OWL Consistency Checker at mindswap
09:45:44<EtnaRosso>i'm checking my ontology
09:46:03<EtnaRosso>i defined a new class <owl:Class rdf:ID="myClass" />
09:46:28<EtnaRosso>but the check result says me
09:46:38<EtnaRosso>Untyped Class: Assuming myOnt:myClass is a class
09:46:50<EtnaRosso>class declaration is wrong?
09:47:38<danja>heh
09:48:41<danja>does that one give you a chunk of rdf/xml to insert for DL? sounds like it prefers it said in a different way
09:51:06<EtnaRosso>i put my rdf/xml into a form
09:52:00<EtnaRosso>wonderweb validator say me as additiona information
09:52:08<EtnaRosso>"Class(a:myClass partial)"
09:52:28<danja>one of the checker things gave suggestions for missing bits - maybe it was the Pellet demo?
09:53:00<Arnia>danja: word of the week: herft
09:53:32<danja>EntaRosso, is it seeing the xml:base ok?
09:53:42<danja>.g herft
09:53:42<phenny>danja: http://en.wikipedia.org/wiki/Roger_Herft
09:53:50<danja>.ety herft
09:53:51<phenny>Can't find the etymology for "herft". Try http://etymonline.com/?search=herft
09:54:01<Arnia>.swhack herft
09:54:41<sbp>that only works on Swhack
09:54:55<Arnia>Oh well
09:55:06<Arnia>phenny: thanks phenny
09:55:07<Arnia>anyway
09:55:21<phenny>No problem.
09:55:24<danja>ACTION started using jbouncer to keep my nick alive - not got around to joining all the usual places (thanks Reto!)
09:58:24<EtnaRosso>danja> EntaRosso, is it seeing the xml:base ok?
09:58:28<EtnaRosso>i can't understand
10:00:49<danja>what URI is rdf:ID="myClass"
10:02:31<rawles>has anyone managed to edit multiple ontologies in Protege-OWL? Or is it just one session, one ontology
10:03:01<EtnaRosso>right danja
10:03:09<EtnaRosso>missing # in a property range
10:03:12<EtnaRosso>thanks!
10:04:26<gromgull>rawles: i think it's only one session per ontology
10:04:28<gromgull>eh
10:04:31<gromgull>vice versa
10:04:39<gromgull>you can import other projects into the current one
10:04:51<gromgull>although no circular imports are possible
10:05:00<danja>rawles, don't think I've tried editing, but the import kind-of worked for me, though it still seemed to balk at some of my non-owlish rdf
10:05:02<rawles>so if i wanted to develop two together i would need to save and open, save and open?
10:05:11<rawles>i really really love protege-owl
10:06:01<rawles>can we make a #protege channel please?
10:08:09<rawles>how do i set the internal name of an ontology?
10:09:41<EtnaRosso>seee xmlns
10:19:58<bijan>rawles, I strongly recommend, if you can, using protege4alpha. It's a ground up rewrite and it's really quite nice. Nice Swoop influence (if I do say myself that a Swoop influence is nice ;)). Doesn't have most plugins, so that's a bear.
10:20:32<Arnia>bijan: is it possible for me to refer to the RDF vocabulary in it?
10:20:53<bijan>You mean as data?
10:20:57<Arnia>I've been trying to get it to import the RDF namespace so that I can use rdf:Statement as a superclass
10:20:59<rawles>can it load prot 3.3 projects?
10:21:09<rawles>i need lots of plugins
10:21:12<bijan>Arnia, ewww
10:21:25<bijan>I wouldn't be surprised if not.
10:21:31<bijan>I might file a bug report if so :)
10:22:09<bijan>To work around that, why not define a "shadow vocabulary", i.e., us the same local names, but make up your own prefix for it
10:22:20<Arnia>heh... that should work
10:22:30<Arnia>I get to be evil in (relative) comfort
10:22:31<bijan>This makes your ontology more portable to reasoners and OWL DL oriented tools
10:22:58<Arnia>well... OWL DL reasoning isn't a... high priority with this
10:23:00<bijan>At the expense of loosing some syntactic facilities in the RDF/XML and not getting special treatement in the triplestore in things like Jena
10:23:02<Arnia>But yeah
10:23:12<Arnia>Hm
10:23:22<Arnia>Actually, no
10:23:24<bijan>That's really the trade off
10:23:31<Arnia>ACTION shakes his head to clear it
10:26:00<bijan>rawles, you are out of luck re: Prot4. Sorry
10:26:00<bijan>Might still be worth developing the owl bits there
10:34:15<rawles>ok thanks
11:14:13<EtnaRosso>i need informations about how to create and use a new type for a typed literal
11:34:15<EtnaRosso>i need to define as property range a closed continuos interval
11:34:45<EtnaRosso>(for example doubles in [0..1])
11:35:08<EtnaRosso>i think the right way is to define a new datatype from double
11:35:25<EtnaRosso>i'm right? i don't know how to
11:43:39<danja>phenny, tell danbri seeAlso: http://www.the-ba.net/the-ba/Events/FestivalofScience/FestivalNews/_Socialnetworking.htm
11:43:41<phenny>danja: I'll pass that on when danbri is around.
12:44:45<m94mni>www.ala.org
13:35:39<EtnaRosso>i defined in owl a class with a cardinality constraint maxCardinality
13:36:07<EtnaRosso>then a new class, subclass of the previous one, with a constraint minCardinality
13:36:27<EtnaRosso>pellet owl validator recognize it as unsatisfiable concept
13:36:33<EtnaRosso>could be a bug?
13:37:02<EtnaRosso>note that maxCardinality and minCardinality has the same value and are about the same property
14:16:48<bijan>EtnaRosso, uhm...it's impossible to tell from your description
14:17:15<bijan>A larger fragment or a pointer to the ontology would be better
14:17:58<bijan>Or if you load it up in owlsight, it will provide an explantion of (in terms of minimum set of axioms sufficient to produce the) unsatisfiability
14:19:06<EtnaRosso>owlsight?
14:19:55<bijan>http://pellet.owldl.com/ontology-browser/
14:20:02<bijan>A:
14:20:15<bijan>A:|OWLsight
14:20:24<bijan>A:A web based frontend to Pellet.
14:20:36<bijan>A:Will show explanations on demand.
14:21:08<bijan>A:Still in early stages; requires a recent browser with GWT happiness. Doesn't support individuals well, yet.
14:21:14<bijan>Enjoy :)
14:33:13<EtnaRosso>nice
15:13:38<bijan>EtnaRosso, did you figure out your issue?
15:16:22<EtnaRosso>no has been helpful to fix other bugs
15:17:01<EtnaRosso>see you
15:31:35<leobard>hi, is there an N3 validator?
15:32:01<x98>any sematnic media wiki people here?
15:48:19<leobard>when I get this at CWM with python 2.5, is it my fault? " from __future__ imports must occur at the beginning of the file"
15:48:38<sbp>nope, bug i' th' code
15:48:52<sbp>easy to fix: just repair what it says
15:48:59<leobard>weird, I downloaded the release.
15:50:37<chimezie>is there a way to get the latest version of CWM w/out getting the entire swap directory from CVS?
15:50:53<chimezie>ACTION would love to bookmark a page with such instructions
15:51:12<sbp>if the filenames haven't changed, you could wget the same structure as is in the released bundle
15:51:23<sbp>if there's no construction for the released bundle, too
15:51:30<chimezie>ugh.
15:51:44<sbp>fine, have fun CVSing! :-)
15:52:24<sbp>the CVS rock vs. the wget hard place
15:53:49<chimezie>the download all of swap via CVS is a non-starter for me, that's why I was asking
15:54:02<chimezie>i'll try wget
16:00:19<Arnia>I'm beginning to get head-spin... a meta-ontology :/
16:25:35<leobard>next problem with cwm: does it have an argument to store the results into a file?
16:27:13<leobard>ok, seems I have to pipe. ok.
17:45:07<timbl>leobard, from __future__ imports must occur at the beginning of the fil
17:45:12<timbl>e
18:12:27<ericP>anyoen seen danbri lately?
18:12:41<ericP>he owes me a posn paper toute de friggin' suite
18:21:45<timbl>Funny, on dbpedia there is a nice data file http://dbpedia.org/data/John_Seely_Brown
18:28:31<timbl>so why does http://dbpedia.org/resource/John_Seely_Brown redirect to http://dbpedia.openlinksw.com:8890/sparql?default-graph-uri=http%3A%2F%2Fdbpedia.org&query=DESCRIBE+%3Chttp%3A%2F%2Fdbpedia.org%2Fresource%2FJohn_Seely_Brown%3E
18:29:35<timbl>which is more of a mouthful an has Content-Length: 3338
18:30:44<timbl>hmmm the sparlw on is in rdf/xml triples
18:31:58<timbl>Same graphs.
18:32:19<timbl>The /data/' file ahs been reserialized .. baked maybe
18:34:44<timbl>Not same graphs. also 2 triples in data, saying... <http://dbpedia.org/data/John_Seely_Brown> :label "RDF description of John Seely Brown";
18:34:45<timbl> <http://xmlns.com/foaf/0.1/primaryTopic> <http://dbpedia.org/resource/John_Seely_Brown> .
18:37:13<Arnia>Oh, they separate uses and mentions then
18:38:16<timbl>Why not just 303 to the /data/ ?
18:39:14<Arnia>Could it be something to do with the way they have the proxying arrangement set up?
18:39:52<sbp>logger: pointer?
18:39:53<logger>See http://chatlogs.planetrdf.com/swig/2007-09-14#T18-39-53
18:40:58<sbp>.email dbpedia-discussion@lists.sourceforge.net 303 See Other for /resource/ URIs? // How about responding 303 See Other from, e.g. http://dbpedia.org/resource/John_Seely_Brown to http://dbpedia.org/data/John_Seely_Brown? See http://chatlogs.planetrdf.com/swig/2007-09-14#T18-39-53 for context. Thanks.
18:41:01<phenny>sbp: Sent (459 bytes).
18:41:11<sbp>there. hopefully it allows posts from non-subscribers
18:54:08<chimezie>ahh the joys of phenny-foo
18:54:54<sbp>I just added calendaring, too, but it only works in The Other Place
18:56:28<chimezie>i still need to port Emeka to the latest module-based framework - so much cleaner
18:56:46<sbp>yeah, it's very handy...
19:20:15<timbl>thanks spb.
19:29:36<sbp>yw
19:38:36<yard>anyone figured out why swig.xmlhack chumps are invisible in Firefox3, Dillo, and WebKitQT? the only time ive ever seen them is loading up firefox2 for release-testing. perhaps this is the infamous 'dont serve xhtml as text/html' problem that Ian Hickson likes to warn about? dillo doesnt even support CSS and theyre invisible
19:40:12<timbl>Ar ethe really not there or just right at the bottom of the lomng page?
19:40:50<timbl>ACTION sighs. .. did some foaf crawling stuff somewhere but don't know where he put it.
19:45:26<yard>yeah div#main isnt even showing up in the DOM inspector. must be a parsing issue
19:45:44<yard>actually it is in webkitQt. just not dillo or trunk firefox (the 2 browsers i use)
19:46:06<yard>anyways theres a better solution - read the RDF version :D...
19:46:57<yard>ACTION wondering about schema for documenting code
19:47:45<yard>methodName, methodFunction, sourceCode, etc
19:48:21<yard>hmm, maybe DC/EARL is enough
19:53:05<kidehen>timbl: http://dbpedia.org/resource/John_Seely_Brown should 303 to http://dbpedia.org/data/John_Seely_Brown
19:53:32<sbp>.head http://dbpedia.org/resource/John_Seely_Brown
19:53:35<phenny>Status: 303 (for more, try ".head uri header")
19:53:36<sbp>.head http://dbpedia.org/resource/John_Seely_Brown Location
19:53:39<phenny>Location: http://dbpedia.org/page/John_Seely_Brown
19:53:47<sbp>/page/?
19:54:10<kidehen>sbp: this is a bug in the proxy setup for sure. It never used to to to /page re. rdf
19:54:25<timbl>You have to ask for RDF
19:54:33<sbp>aha
19:54:35<timbl>$ curl -I -H 'accept: application/rdf+xml' http://dbpedia.org/resource/John_Seely_Brown
19:54:35<timbl>HTTP/1.1 303 See Other
19:54:35<timbl>Location: http://dbpedia.openlinksw.com:8890/sparql?default-graph-uri=http%3A%2F%2Fdbpedia.org&query=DESCRIBE+%3Chttp%3A%2F%2Fdbpedia.org%2Fresource%2FJohn_Seely_Brown%3E
19:54:59<timbl>kidehen, is the /data/ file baked?
19:55:10<timbl>(or is it a proxy to the sparql?)
19:55:47<kidehen>timbl: in both cases (before i.e. /data) and what appears to be the case now, it's based on 303 proxy services from (pubby or virtuoso)
19:56:12<kidehen>timbl: something appears to be going on as the 303 to the actual sparql endpoint is a new development
19:56:32<kidehen>timbl: i.e. appears to be happening as we type or something :-)
19:56:33<timbl>I thought it was always like that, and 'data/ was a new development!
19:56:54<kidehen>timbl: pubby always routed to /data
19:57:21<timbl>I prefer the 303 t the dat file, it just leaves a better URI for people to bookmark if they want.
19:57:22<yard>dajobe: the <script /> tags break dillo and firefox3. needs to be <script></script> for whatever reason
19:57:30<yard>dajobe: in swig.xmlhack
19:57:38<kidehen>timbl: anyway, this is all happening outside our domain (i.e. in Germany)
19:57:40<yard>dajobe: everything past that script tag doesnt make it into the DOM
19:58:31<kidehen>timbl: we are setting up another server so that we can demonstrate # URIs as part of the current discussion re. DBpedia URIs
19:58:49<timbl>That will be interesting.
19:59:23<kidehen>timbl: so there will be another server in parallel so that we can work our way towards what works best for everyone
19:59:43<kidehen>timbl: two live servers will enable the pros and cons to be evaluated by all interested parties
19:59:48<timbl>Currently, we are having a real problem with the 303s in Ffox. Firefix insists on alwats making sure the URI bar address what is on the screen, and it insists on folloiwng redirections.
20:00:08<kidehen>timbl: personally, I prefer # URIs :-)
20:00:18<timbl>so something like /data/John_Seely_Brown#this ?
20:00:24<kidehen>timbl: yep!
20:01:29<dajobe>yard: ?
20:02:25<timbl>That will be interesting. It will work very nicely with the tabulator.
20:02:59<timbl>Coul change the default label generator so that if it find #this is uses the bit before as a label, not the 'this'.
20:03:07<yard>dajobe: apparently, serving xhtml as text/html is part of the problem. since it ignores the /s at the end of tags. then barfs when script doesnt have a closing tag
20:03:27<yard>dajobe: at least anne van kesterns blog says that
20:06:00<dajobe>ok, I added a space inside so xslt won't generate the empty form tag
20:06:05<dajobe>maybe I should just switch the miem type
20:08:01<c>cool, fixed :)
20:10:53<c>changing MIME might introduce more probs. i cant read Shelley Power's site cuz of some parser error..
20:10:53<dajobe>ok, swig.x.c is now application/xhtml+xml mime type
20:10:56<dajobe>good luck browser
20:11:06<dajobe>heh, out of sync irc
20:11:32<dajobe>there should be no parser errors, it's generated from valid xml
20:11:46<c>yep, it still works in ff3,
20:11:52<c>but now Dillo tries saving swig.xmlhack to disk
20:11:56<dajobe>I meant more about ie
20:11:57<c>ah well. dillo is only for emergengies
20:12:00<dajobe>well dillo is just dumb
20:12:22<dajobe>I don't use IE\d but some people do, does it handle xhtml mime types or not
20:14:51<dajobe>ok, lynx doesn't handle it either, I'm reverting
20:14:59<dajobe>at least until I do some kind of Accept: checking
20:15:13<kidehen>timbl: another option I am considering is adding the ability to produce "xyz.rdf" for SPARQL protocol URLs as part of the Linked Data Deployment features of Virtuoso. This is functionality we always had re. exposing anything in Virtuoso as a WebDAV resource
20:15:44<c>wikipedia IE page says it treates xhtml+xml as 'file for download' as well
20:15:50<c>maybe fixed in v7?
20:16:39<kidehen>timbl: so you have a traditional web information resource url returned in "Location: .."
20:16:51<kidehen>timbl: for "xyz.rdf"
20:17:47<timbl>WebDAV?
20:17:50<kidehen>timbl: I think this option is mandatory for reducing the activation threshold of linked data deployment as currently espoused across the LOD community
20:17:54<kidehen>timbl: yes
20:18:31<timbl>How does webDav fit in ... not webdav properties.... webdav authoring?
20:19:33<c>dajobe: with CSS float:left on #main and float:right on #sidebar, could eliminate need to scroll past 3 pages of links in non CSS browser
20:19:45<c>dajobe: since presumably you want the new stuff on top, insteazd of bottom
20:20:33<kidehen>timbl: we use WebDAV to expose resources to HTTP, so in this context I mean: to the UA its an HTTP server interaction, but inside Virtuoso we are creating a handle via the WebDAV functionality realm. Of course this same resource is accessible via WebDAV too in the normal sense. And the WebDAV resource properties also synced with the RDF Store (again, all of this happens inside Virtuoso)
20:22:14<kidehen>timbl: so Virtuoso will look like Apache re. the resource interaction, but go beyond Apache in the sense that it's producing the resource handles on the fly based on SPARQL protocol URLs. The user simply tells Virtuoso how he/she wants to name the resource etc..
20:22:16<dajobe>c: i've changed the css, does that help?
20:22:42<c>dajobe: well, #main would have to be above #sidebar, in the XML as well..
20:23:06<c>dajobe: siblings of the same parent element, i mean, but order reversed
20:23:14<dajobe>try now
20:23:21<c>awesome
20:23:26<c>in lynx..
20:23:37<c>ty
20:24:39<c>its pushing the sidebar down though. clear:none missing somewhere (or more likely something set to display:block)
20:24:43<kidehen>timbl: re. authoring, you wouldn't be able to edit said resource unless we have a setting that tells Virtuoso to persist the resource rather that generate on the fly. This is an option that I will have the guys add to this feature
20:24:44<c>i should tweak the css and send you a patch
20:25:25<kidehen>timbl: but said resource will be out of sync with the original as it is now a completely separate thing :-)
20:25:48<dajobe>hmm, that looks bad on safari
20:26:12<dajobe>when it's too narrow
20:26:42<c>yep, floats are supposed to coexist in the same horizontal space. so theres something clearing it
20:27:19<dajobe>ACTION mtg
20:40:15<c>dajobe: the sidebar is appearing _inside_ the last item element, perhaps this is <a name=''/>s not being closed or something..
20:41:41<carmen>ACTION wonders if FF parser can run in 'debug mode' with verbose
20:43:31<carmen>dajobe: not to mention the 2nd item element is now child lf the first item el, etc :/
20:57:33<carmen>dajobe: <div class="comments"/> causes that extra nesting, which i guess doesnt happen if you serve as xhtml+xml, but then IE/lynx choke so... pick your poison?
20:58:00<carmen>dajobe: final note, i think you wanted <A id='linkid234234'/><a href=http://chumpedlink>name</a> to actually be one element?
21:01:57<carmen>at least links are visible now - major improvement
21:25:00<dajobe>can you send this by email, I'm busy
21:42:49<carmen>dajobe: so in summary - not outputting div.contents when theyre empty, and setting an explicit max-width on div.content is looking like the best solution, since the other option (moving content below sidebar) is awful in lynx/non-css and wrapping both main and content in a styling div is weak
21:43:22<carmen>maybe like max-width:75% or smth..
22:59:04<Arnia>Right, completed the first draft of my non-axiomatic logic ontology
22:59:22<Arnia>yvesr: let me know if you want to take a look

Back to channel and daily index: content-negotiated html turtle