Experimental IRC log swig-2008-02-18

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.

00:12:16<nslater>would it be easy to make a grddl profile for docbook->foaf... im a grddl n00b
03:59:18<mhausenblas>ACTION wave hello to the wonderful SWorld
03:59:29<nslater>ACTION waves back
04:00:29<si>greetings
04:01:13<mhausenblas>hi there
04:01:28<mhausenblas>so, we finally entered LAst Call with RDFa :)
04:01:36<mhausenblas>http://www.w3.org/MarkUp/Drafts/ED-rdfa-syntax-20080217/Overview.html
04:01:57<mhausenblas>A:| Last Call Working Draft of RDFa Syntax
04:02:17<mhausenblas>ACTION hopes we make it ;)
05:03:00<adu>what does it all mean?
05:04:25<si>42
05:04:42<adu>thank you! finally an answer!
05:04:55<adu>now, what was the question again?
05:04:55<adu>lol
05:05:23<adu>I'm wondering what the problem is with technology
05:05:34<adu>there is certainly enough of it
05:05:45<adu>the problem isn't a lack of technology
05:05:54<adu>the problem isn't a lack of solutions
05:06:14<si>what is lacking a solution?
05:06:18<adu>the problem is immediacy
05:06:27<si>i think part of the porblem is what is 'profitable' is not always what is most useful in the long run for everyone
05:06:40<adu>that too
05:07:38<si>so what gets invested in is pretty predicatble non-novel and useless (ie, facebook apps, with some ad-monetization MBA promise somewehr)
05:08:00<adu>whats MBA?
05:08:07<si>Master of Business Administration
05:08:09<adu>o
05:08:13<si>its a type of degree offered by unviersities
05:08:16<adu>ic
05:08:45<si>metaweb got 60 million to make a proprietary version of essentialy Virtuoso or Sesame (god knows what their business plan is)
05:08:52<adu>i don't see how that has anything to do with 'ad-monetization'
05:08:54<si>1/100th of that could keep like 6 people working on sesame or virtuoso for a year..
05:09:10<si>but of course, then Goldman wouldnt be able to flip their MWEB shares to Yhoo at a 200% profit, so it doesnt work that way
05:09:32<si>considering MySQL got bought for so much, maybe things will change. but i dont have my hopes up. MYSQL and SKYPE were pretty rare (and one might argue idiotic) deals
05:09:40<si>what solutions do you need?
05:09:42<si>that dont exist
05:09:45<si>or are not good enough
05:09:51<si>or whatever was your question/probilem
05:10:45<si>ive got a whole laundry list
05:10:53<si>im currently on 'web browser that doesnt suck'
05:10:58<adu>my current problem has to do with DHCP/NAT/PAT but thats a different story
05:11:00<si>dillo2 is good but has bugs. so i think im gonna write one using GTK2HS
05:11:07<si>beause i cant really hack on c++ to save my life
05:11:23<si>oh yeah
05:11:31<si>we need IPV6 and HIP and the death of NAT, definitely
05:11:33<adu>I love Haskell
05:11:42<adu>whasts HIP:?
05:11:57<si>decouples IP from its host identity function
05:12:07<si>IP is location, HIP is a persistent identifier that outlives address
05:12:24<si>so you can do distribyted semweb stuff (assuming NAT isnt in the say) and have it persist over wifi changeups, etc
05:12:34<si>i change my IP address a ton, and everything breaks :/
05:12:37<si>irc included of course
05:12:39<adu>o wow, that would solve so many problems!
05:12:44<si>but also smtp tunnels to my mutt server, etc
05:12:47<adu>not to mention dynamic IP
05:12:52<si>yes
05:12:57<si>HIP can 'renegotiate' without closing conection!
05:13:02<si>persistence across IP change
05:13:10<si>if both IPs change at once, it can use DHT to lookup the new IP
05:13:10<adu>that would rock!
05:13:20<si>theres lots of IETF drafts and a couple implementations
05:13:28<adu>but all that stuff should just work
05:13:32<si>yep
05:13:34<si>nobody is funding it
05:13:36<si>since its not a facebook app
05:13:40<si>or a potential GOOG acquisition target
05:13:48<si>if we had distribute p2p semweb, nobody would need google anymore
05:13:54<adu>I'm always (being for the last 10 years), about semantics and representations
05:13:57<si>they couold just find what they wanted via faceted searching of aDHT or whatever
05:14:02<si>kind of like torrents already are ;)
05:14:10<si>content piracy is the leader of new technology
05:14:50<si>so, i guess my plan is create a big enough competitor to google, that i get acquired or they try to kill me (with money)
05:15:04<nelix_>si: wb that does not suck is high on my list
05:15:06<si>since i cant get people to hire me to make "DHT enabled HIP-URI based semantic platforms' or ahtever
05:15:08<nelix_>si: although kde is pretty good
05:15:10<si>wb?
05:15:14<nelix_>web browser
05:15:20<si>konqueror?
05:15:23<nelix_>yeah
05:15:24<si>firefox is just too slow on linux
05:15:31<si>and too much kitchen sink
05:15:33<nelix_>agreed
05:15:37<si>esp now wiht SQL, and offline stuff, and this and that
05:15:37<nelix_>k is just as bad
05:15:40<nelix_>although much faster
05:16:14<adu>I've been designing an IDE for the last 10 years, and so far my plans involve the following projects: http://interreality.org/ (also #vos), almost all W3C specs, Haskell, and Parrot
05:16:21<si>with real semantic data formats, we dont depend on HTML as much
05:16:26<si>or custom JS
05:16:30<si>you can just make a ncurses rdf browser
05:16:34<si>and some input filters in GRDDL or otherwise
05:16:39<si>and browser in the terminal, gopher style :)
05:16:47<si>or of course, browse in something thats not slow as hell like firefox
05:16:55<si>some minimal cairo/pango thing or qt thing or whatever
05:17:09<si>cool
05:17:19<si>did you recently announce your IDE to haskell list, or was that something else
05:17:24<si>i tried Yi, it used incredible RAM. back to emacs ;0
05:17:25<adu>nope
05:18:02<adu>si: I have no code, unless you count the hundreds of 10K projects scattered across my HD amounting to failed attempts at code management
05:18:11<si>hee
05:18:16<si>i started making git repos all over my drive
05:18:17<si>for that purpose
05:18:22<adu>heh
05:18:42<si>when its good enough i push to an externa USB HD, or setup acct on repo.or.cz...
05:18:51<si>but yeah, i understand completely
05:18:58<adu>I have 2 projects that have come pretty far, but they are just opengl apps written in C and Python
05:18:59<si>to me, refactoring and rewriting is important
05:19:08<adu>now I'm working on an OpenGL shell in Haskell
05:19:11<si>cool
05:19:23<adu>I think this will be the one...
05:19:30<si>i am mainly in search of decent canvas classes atm
05:19:33<adu>Haskell is so easy when it comes to code management
05:19:39<si>QT is nice, but theres no haskell/ocaml bindings afaik
05:19:56<si>Cairo afaik does not even have a 'bind events to drawn pieces' api
05:20:03<si>you have to build your own canvas class on it..
05:20:28<adu>hmm
05:20:31<si>so theres a few (gocanvas, etc) bt all grossly immature and mildly broken
05:20:40<nelix_>si: if i were to make a WB, i would target xhtml2 rendering, with no css support at all
05:20:41<si>anwyays, im going to try gtk2hs, and see how it is
05:20:50<si>nelix_: how would styling work?
05:20:56<si>nelix_: no styling?
05:20:56<nelix_>it wouldnt
05:20:59<si>heh
05:21:11<nelix_>i like information
05:21:13<si>same
05:21:23<nelix_>art is cool, but i need not have my information woven into complicated art
05:21:29<si> i disable CSS and JS in firefox
05:21:43<nelix_>you might have some basic user css
05:21:47<si>well
05:21:49<si>i use the layout
05:21:51<si>but not the colors
05:21:57<adu>I use NoScript
05:22:00<si>same
05:22:05<nelix_>the WB stack is so huge
05:22:12<nelix_>so many standards and half standards
05:22:14<si>with many interlocking dependencies
05:22:20<si>have you read HTML5?
05:22:23<si>its a beast
05:22:44<nelix_>yeah
05:22:48<nelix_>worst idea ever
05:22:51<adu>have you heard of VOS?
05:22:54<nelix_>no
05:22:58<si>what is that?
05:23:38<adu>it stands for Virtual Object System (http://interreality.org/)
05:23:41<adu>also #vos
05:23:42<si>oh
05:23:51<si>like OS2 network object system?
05:24:03<adu>its the missing piece to VRML/X3D
05:24:14<si>this looks like Croquet
05:24:35<adu>its similar in theory
05:24:54<adu>although Croquet was gear to be 3D from the beginning
05:25:16<adu>VOS provides much more low-level features, and adds 3D as an add-on
05:25:21<si>are the hyperlinks in VOS compatible with HTTP?
05:25:37<adu>um, i dunno
05:27:17<adu>but anyways, I'd like to model the interface of my IDE after either: http://www.perceptivepixel.com/ or: http://bumptop.com/
05:27:39<si>id like an IDE to work with networked resources
05:27:45<si>and local
05:27:51<adu>si: what do you mean?
05:27:58<si>accesing local files in firefox is so suboptimal/difficult
05:28:38<adu>resources meaning preferences/default-skin?
05:28:50<si>i mean instead of writing code youre writing apps using a GUI
05:29:03<adu>Visual Programming?
05:29:08<si>and its all running locally, using HTTP/RDF as needed (and has a HTTP server of its own running)
05:29:13<adu>or Tangible Values?
05:29:34<si>ive used Pure Data in the past, and made some GUI extensions. and have been reading on conal Elliot's stuff
05:29:52<adu>http://conal.net/papers/Eros/
05:29:53<si>but im coming at this from music composition, and networked-app collaboration tools
05:30:02<si>rather than 'synthetic image filters' which seems to be his main test case
05:30:18<si>he's stil god tho ;)
05:30:31<si>im not really sure how much 'difficult' stuff he glazes over, in terms of whats possible in pure functional
05:30:34<adu>who is?
05:30:43<si>like i guess gtk2hs people cheat and store state in GTK widgets, and what not
05:31:08<si>im happy storing state on the local FS, in a RDF store
05:31:30<adu>ya
05:31:38<adu>o you were talking about Conal Elliot
05:31:40<adu>ic
05:31:46<si>i havent read all the reactive papers yet
05:31:54<si>im not sure how much of that is useful for an IDE
05:32:08<si>where youre more manipulating source or some kind of visual AST
05:32:43<adu>Ya, effective real-time tree manipulation is essential
05:33:01<si>didnt Apple already clone most of jeff han's stuff on iphone
05:33:09<adu>since tree's appear in XML, RDF, FileSystems, Scene Graphs, GUIs
05:33:10<si>i mean, its just finger gestures generating differnet input, no?
05:33:17<si>he doenst go into 'workflow' much
05:33:41<adu>nope, Jeff Han's Gestures are quite different
05:33:52<si>trees appear in RDF? you mean the implicit tree in the URI paths?
05:34:02<adu>no, i mean graphs
05:34:04<si>some people say URIs are opaque, and sure they are. but that / is useful
05:34:10<adu>tree's are special cases of graphs
05:34:16<si>ok
05:34:24<si>acyclic
05:34:36<adu>right
05:34:58<si>but yeah thats why im into haskell now. you cant really rebalance a big tree in almost- realtime in python/ruby
05:35:00<adu>but I think the key-combo is already decided for trees in general
05:35:07<si>w/o some orthogonal C lib anyways
05:35:15<si>which is?
05:35:19<adu>my favorite way of navigating trees is the MacOSX Finder in column view
05:35:27<si>mm
05:35:34<si>how do you switch columns
05:35:36<si>ctrl-right?
05:35:39<si>tab?
05:35:44<adu>left/right/up/down/first-char
05:35:49<si>i havent used OSX since 10.0 :)
05:35:49<adu>oh, <Command>-3
05:36:19<adu>no, to switch collumns you just press left/right
05:36:27<si>ok
05:36:36<adu>up/down moves the selection to different files in that folder
05:37:09<si>surely you have some ideas on 3d tree manipulation
05:37:10<adu>thats the way navigating trees should be
05:37:17<si>i dug 'greg's browser' as it was called on os8
05:37:25<si>which was just the nextstep browser rewritten for Mac
05:37:27<adu>well, graphs are much different
05:37:49<si> always laugh at that isaviz stuff
05:37:52<si>who the fuck would acutally use that
05:37:55<si>nobody does
05:38:08<si>its just to get grant funding ('oh wow, gee wiz bang, you guys rule'
05:38:15<si>you animate the arcs moving around hehe
05:38:34<adu>does that use graphviz?
05:38:45<si>no but same idea
05:38:48<adu>ic
05:39:01<si>you show the current resource and all the arcs to some depth, i guess
05:39:05<adu>well, graphviz would be nice if it was real-time, but its more of a compiler now
05:39:25<si>i dont mind using 'familiar' interfaces, but faceted filtering is so common that a generic one that could be used for email/blgoposts/whateverelse is fine to
05:39:32<si>yah
05:39:42<si>if there was a graphviz class for qt, or something
05:40:04<adu>Apple had a prototype for viewing trees in 3D called 'HotSpot', but I didn't like that very much
05:40:13<si>hrm, tihnk i tried it
05:40:17<si>or one called 'Project X'
05:40:20<si>which was like a 3d version of cyberdog
05:40:22<si>if ir ecall
05:41:26<si>theres not even ver many 2d browsers
05:41:28<adu>heh CyberDog was a joke
05:41:58<si>the barriers of entry are so high, since people like their JS and CSS and layout perfection so much
05:42:30<si>if i had money id fund at least a half dozen web browsers. dillo to start with, btu also lightweight classes for qt, gtk..
05:43:08<si>dillo2 is fast as hell. seems to be freezing up on blocking HTTP reqs for cookies, tho :/
05:43:10<adu>Ideally, the only system interface for trees would be the scrollbar, which would serve 2 purposes, and the elements of a tree or a graph would be displayed according to some plugin-system where elements could decide their own rendering if it wasn't the default rendering.
05:43:29<si>just a sroll bar?
05:43:41<si>lke for siblings on one node?
05:43:48<si>a 3d scroll handle?
05:43:48<adu>(1) Scrollbars would indicate the connection between the parent/self and the children/links-from-self, and
05:44:18<adu>(2) Scrollbars would allow the user to skim thru a large amount of chilren/links-from-self
05:44:53<adu>my idea isn't really 2D/3D since I have instances of it imagined for both
05:45:39<adu>if you look at MacOSX Finder column view the only thing between columns is a scollbar, so it makes sense
05:45:59<si>so scroller + arrows?
05:46:11<si>its important to be easily usable from _only_ keyboard or _only_ mouse
05:46:15<si>mac likes to require keyboard _and_ mouse :/
05:46:20<si>and htey call themselves HI gods heh
05:46:25<adu>yes, but the current idea of a scrollbar needs serious rethinking
05:46:35<si>i never use them
05:46:37<si>i use pageup/down keys
05:46:40<si>arrows for finer grain
05:46:52<si>the fact you have to 'fish them out' sucks for one
05:46:55<si>so i think jeff han has the righ tidea
05:46:57<si>screw em
05:47:04<si>scrolling itself doesnt need to go away
05:47:05<si>thats good
05:47:07<si>its just the bars
05:47:21<si>they waste space and are finicky
05:47:38<adu>well, the 3D Scollbar could be simply a set of lines from the parent-node to the child-nodes
05:47:38<si>plus they have an 'end'
05:47:47<si>most of my stuff doesnt end, if anbything youd eventyaly loop back
05:47:49<adu>there wouldn't have to be any arrows anywhere
05:48:15<si>so you cycle thru children on a right motion or something?
05:48:28<si>i mean, you have tos omehow say 'go down in the tree' vs 'iterate siblings
05:48:53<adu>what do you mean "iterate siblings"?
05:49:04<si>arcs sharing a common parent, or whatever
05:49:20<si>at this point, i have a ton of ideas, and could read a ton more, and need kickass canvas/scenegraph classes
05:49:27<si>pref ones that work from haskell ocaml or scheme
05:49:36<si>OSG is like.. big. and stuff.
05:49:41<si>Cairo lacks even 'what did i just click on' functionality
05:49:50<si>and QT is a monolithic hog (but it does have a nice canvas class)
05:49:56<adu>oh, I though I described that already by how MacOSX Finder column view does it... up/down selects siblings and left/right selects parent/first-child
05:50:01<si>i think gentoo just broke up qt in 4.4, nm
05:50:12<si>well
05:50:13<si>when querying
05:50:18<si>you are talking about merging 2 trees
05:50:21<si>set intersections
05:50:34<si>i guess you could just use a column view for each facet
05:50:37<si>but, theres other poossibilities too
05:51:01<si>i wnt an 'exhibit' that works on millions of nodes in the dataset
05:51:10<si>which means, for one dont write it in JS, in a browser
05:51:18<si>but also , theres a ton of GUI ideas that havent been tried
05:51:59<si>newegg is good
05:52:03<si>i always can find what i want real fast
05:52:14<si>eg a '3.5 " HD > 800 gb made by samsung or seagate and costing less than 300" etc
05:52:29<si>usualyl its just a single facet is narrows at once, and the possible facets are presented
05:52:50<adu>what do you mean by "exhibit"?
05:52:59<si> http://simile.mit.edu/exhibit
05:54:31<adu>hmm
05:54:58<si>i store my RDF in the fs tree, ive browsed it with Konqueror before
05:55:04<si>but kdelibs is such a bloat
05:55:09<si>and id have to make some konqueror plugins
05:55:15<si>not sure its worth it reinstalling KDE just for a halfway decent tree-browser
05:55:23<si>is there a osx style colun browser for linux
05:55:26<adu>have you heard of CyN?
05:55:26<si>i think theres 'evidence' for e17
05:55:29<si>nope
05:55:46<si>i guess i could use FUSE
05:55:53<si>to remount the existing fs as a virtual one
05:55:59<si>where the metafiles are made invisible
05:56:07<si>and then use existing tree browsers...
05:56:19<si>thing is , all the fs browsers make assumptions
05:56:26<si>like youd only one one reosurce in the preview pane at once
05:56:32<si>not say 10 , or 20
05:56:33<si>eg a resultset
05:56:41<si>all being previewed (eg their text contents, image contents, etc)
05:56:52<si>so, im not sure even trying to force things to work in eisting browsers is worth it
05:57:07<si>i guess theres 'icon view', but good luck maing a text literal display readably in that
05:58:27<adu>I've been learning alot about patterns lately
05:58:52<si>got any reading suggested?
05:58:56<si>im more on an implementation kick
05:59:04<adu>well
05:59:12<si>reading Okasaki's DS book. trying to find a canvas/scenegraph class thats minimalist and doesnt suck, etc
05:59:16<adu>I have some languages to learn
05:59:31<si>more and more thinking of bilding stuff on tinyscheme or similar. or hacking up
05:59:53<si>Fluxus heh
05:59:56<si>ghc is just enormous, im fine with a less complicated type system and list comprehension syntax if it means a smaller toolchain
06:00:08<si>whats that gradual-typed scheme i saw mentioned on arc forum, hmm
06:04:08<adu>Regex, Awk, Haskell "case", Scheme "syntax-rules", M4, and http://web.engr.oregonstate.edu/~erwig/papers/abstracts.html#Haskell00
06:04:49<adu>I love the file name "PGandTP" hehe
06:06:21<adu>http://reference.wolfram.com/mathematica/guide/Patterns.html
06:10:09<adu>http://www.daimi.au.dk/~beta/Manuals/latest/beta-intro/Patterns.html#H1_3
06:12:03<si>thx
06:12:16<adu>http://research.microsoft.com/~simonpj/Haskell/guards.html
06:12:32<adu>i love patterns
06:13:52<si>heh
06:13:56<si>i got ~erwig in my papers queue already
06:14:06<adu>papers queue?
06:14:15<si>directory of ps.gz and pdf files to read
06:14:19<adu>ic
06:14:22<si>i havent buildt a mind-uploading machine yet
06:14:30<adu>lolol
06:14:36<dorian>when you do i'd like one
06:14:39<adu>me2
06:15:00<dorian>i'm trying to get a grand unified theory together
06:15:11<dorian>of how to get stuff into my brain en masse
06:15:14<adu>i have one, but it explains nothing
06:15:46<adu>o learning? my mom is an expert at learning
06:16:04<dorian>mostly a matter of how to organize the time
06:16:18<dorian>i'm trying to do a self-similar model of time management
06:17:08<dorian>grand supercycle sort of thing
06:18:12<adu>(1) Plan (2) Monitor (3) Evaluate (4) Customize
06:18:23<adu>the four learning strategies
06:18:44<adu>my mom's written many books about them
06:19:33<si>im not a robot. learning comes automatic whne theres curiosity
06:19:38<si>maybe 1.2.3.4 stuff wroks for robots
06:19:49<adu>heh
06:19:56<si>'learning for dummies'
06:20:00<si>she should write that one
06:20:07<adu>indeed
06:20:08<dorian>i got a copy of 'how to solve it' by george polya
06:20:13<dorian>it's pretty interesting
06:22:33<adu>From my experience, solving generalizations is easier than solving specializations
06:23:53<dorian>there's a lot more entropy in the latter
06:24:16<adu>whats entropy?
06:24:34<dorian>randomness
06:24:56<adu>how is solving special problems random?
06:25:01<adu>i just don't see it
06:25:05<dorian>nevermind.
06:29:38<adu>si: do you have any papers about Join calculus?
06:29:48<si>nop
06:30:08<si>hrm
06:30:12<si>sounds like the pi calculus a bit
06:30:17<si>i think maybe i read hte wikipedia page :)
06:30:23<adu>yes, for concurrent computation models
06:30:29<adu>also http://en.wikipedia.org/wiki/Unification
06:30:33<adu>for logic programming
06:30:54<adu>very similar to pattern matching
06:39:16<adu>Well, I'm getting tired, I might as well try to pattern match the pillow on my bed with my head, unless unification fails... :P
08:25:44<chlorium>hey, i'm experiencing some problems working with swig, c++, and python
08:25:53<chlorium>i have a method that returns an std::list, like so:
08:26:00<kwijibo>wrong swig :)
08:26:09<chlorium>er
08:26:11<chlorium>oh
08:26:12<chlorium>heh
08:26:19<chlorium>i usually ignore the topic
08:26:28<chlorium>but that's not such a good policy :-)
08:26:49<kwijibo>no worries
09:18:00<EtnaRosso>morning all
09:19:30<kwijibo>morning
09:27:28<tommorris>http://barcampbrighton.eventwax.com/barcampbrighton2
09:27:37<tommorris>F:|BarCamp Brighton 2 signup
09:27:53<EtnaRosso>danja are you there?
09:28:00<tommorris>F: At 1230 GMT, get over on to this page and signup!
09:28:45<tommorris>F: /me hopes that SemanticCamp might prompt more SWIGgers to go to BarCamps.
09:42:09<libby>F: the london one was good fun - thanks tommorris and all :-)
09:42:28<kwijibo>ACTION waves to tommorris and libby
09:42:41<kwijibo>yes, cheers tommorris
09:47:37<tommorris>thanks folks
10:01:51<EtnaRosso>i'm searching some example of "weighted" roles implementations using DL
10:02:26<EtnaRosso>for example how to encode an assertion like
10:02:41<EtnaRosso>"The distance between A and B is 4 Miles"?
10:05:38<gromgull>http://www.semanticscripting.org/SFSW2008/challenge.htm
10:05:51<gromgull>G:|Semantic Scripting Challenge at ESWC2008
10:06:15<gromgull>G:Part of the Scripting for Semantic Web workshop: http://www.semanticscripting.org/SFSW2008
10:06:30<kwijibo>EtnaRosso: would that be an n-ary relationship? http://www.w3.org/TR/2004/WD-swbp-n-aryRelations-20040721/
10:06:30<gromgull>G: Wonderful amazing prizes from Talis!
10:06:42<Arnia>Hm. Interesting
10:07:03<EtnaRosso>really i was thinking something quite different and more processable
10:08:19<EtnaRosso>to define classes like
10:08:45<EtnaRosso>"less or equal 13 Miles" and put
10:09:25<EtnaRosso>and individual (A,B) that belongs to this class
10:09:32<EtnaRosso>i'm not shure it is a good idea
10:12:25<gromgull>G:First prize kindly sponsored by Talis!
10:33:55<tommorris>danja_: dannyayers.com seems to be down
10:34:21<danja_>thanks tommorris - currently rebooting
10:34:53<danja_>ACTION wonders if tommorris has blogged re. camp yet
10:35:04<tommorris>I'm just writing up a guide on how complex microformats could use a GRDDL-like process for complex meaning determination and wanted to point to your discussion of GRDDL and "following one's nose"
10:35:19<tommorris>It's a discussion we had at SemanticCamp
10:35:35<tommorris>I'll meta-blog it later - ie. talk about the camp rather than about some of the ideas
10:37:50<danja_>cool
10:38:23<danja_>ACTION points to timbl & DanC re. nose-following
10:38:51<danja_>.g the basic follow the nose way the web works
10:38:54<phenny>danja_: http://inkdroid.org/journal/2008/01/04/following-your-nose-to-the-web-of-data/
10:39:41<tommorris>thank you. gotta go out but will finish it later
10:39:51<danja_>cheers tommorris
10:39:57<tommorris>see http://premasagar.com/events/semanticcamplondon/demo for the sort of HTML I'm trying to figure out
10:40:15<danja_>.g the basic follow-your-nose way the web works
10:40:18<phenny>danja_: http://blogs.talis.com/mt/mt-search.cgi?IncludeBlogs=7&search=rules&submit.x=1
10:40:24<danja_>heh
10:41:04<danja_>sometimes you need a long nose
10:41:57<danja_>http://www.w3.org/DesignIssues/diagrams/arch/follow.png
10:42:18<danja_>H:| the basic follow-your-nose way the web works
10:43:05<danja_>H: timbl's diagram - fun already, can't wait for it to get grddl & sparql too
10:59:25<si>ACTION finds hypertext.rb in ruby-gnome2 and proceeds with firefox elimination plan
11:07:46<tuukkah>one more complication to the diagram: rdf autodiscovery links from html
11:08:10<tuukkah> rdfa
11:31:40<kwijibo>tommorris: how did your rdf for beginners session go?
11:44:01<bengee>http://groups.drupal.org/node/8930
11:44:03<phenny>bengee: 15 Feb 20:02Z <mhausenblas> tell bengee /me looking at http://drupal.org/project/arc_rdf_store and dunno where to download the module or am I kinda stupid?
11:44:41<bengee>I:|RDF API for Drupal 6.x
11:48:25<bengee>I: geting closer to a coordinated effort for RDF in Drupal
11:48:37<bengee>I1: getting closer to a coordinated effort for RDF in Drupal
12:02:41<kwijibo>ACTION liking the api
12:13:47<mhausenblas>hey bengee
12:14:30<bengee>heya
12:14:49<mhausenblas>hm, phenny did warn you already, right ;)
12:15:11<bengee>no idea about the download, the project is maintained by scor
12:15:26<mhausenblas>ah. ok. so I need to contact him directly?
12:16:22<bengee>yeah, probably best
12:16:28<mhausenblas>ACTION dunno, but RDF inport is possible with that module, right?
12:17:03<mhausenblas>ACTION lets user populate end-user with Protege and then wants to sump it into Drupal
12:17:24<mhausenblas>ACTION thinks he can not write properly any more ;)
12:17:27<bengee>I'd think so, but the module is apparently still very early
12:17:44<mhausenblas>yeah yeah I know, king-o-modest :)
12:18:15<mhausenblas>ACTION should not be so cheeky
12:19:06<mhausenblas>ACTION still thinks that this would be a good showcase for Drupal/RDF, as we have a customer who is willing to give it a try
12:19:33<mhausenblas>ACTION knows that we are willing to sell it :)
12:20:01<mhausenblas>ACTION thanks bengee and keeps him posted as we progress. biab
12:23:23<bengee>great
12:47:22<tommorris>kwijibo: The RDF for beginners session went pretty well. Might be an idea if we had a document explaining the best approach to teaching complete beginners about RDF
12:47:50<kwijibo>so who went, what did you say, what feedback did you get ? :)
12:48:24<tommorris>so, on sat night a fair few people asked over dinner saying that they found the talks interesting but too high-level for people who don't know about RDF
12:48:41<tommorris>ACTION pulls up slides with intention to upload
12:48:45<kwijibo>s/high/low ?
12:49:03<tommorris>high-level - ie. they presumed knowledge of RDF
12:49:37<kwijibo>ah, high, as in hard, not as in overview
12:50:01<tommorris>in the session, i tried to explain that RDF was a model not a syntax
12:50:10<tommorris>and used N3 as the primary way of teaching
12:51:16<kwijibo>did folks grok it?
12:51:22<tommorris>I pegged it to http://www.w3.org/2000/10/swap/Primer.html (timbl's N3 primer)
12:51:24<tommorris>yeah, I think so
12:51:47<tommorris>the problem is that if you give a "here's the sort of thing you can do with it" talk, everyone thinks "yeah, but I can do that some other way"
12:52:03<kwijibo>yeah
12:52:09<kwijibo>exactly
12:52:13<tommorris>and if you give a "here's the abstract model, and, hey, there are tools" talk, everyone thinks "this is interesting, but I don't know how to do anything"
12:53:16<tommorris>I did show off rdflib in Python and also Tabulator in Firefox
12:53:28<kwijibo>pretty much any 'one' thing that you can do with rdf, you can do with something else - it's the multitude of different things you can do that makes it special
12:54:18<tommorris>exactly
12:54:28<tommorris>and you sort of have to take it on faith that there's an interesting destination
12:54:44<kwijibo>eg: you can do much the same with rss2 as sioc - but using sioc you can enable lots of different other things in the same data
12:55:27<tommorris>some people tend to think of RDF as being "sorta like XML", some think it's "sorta like SQL" and some think it's "sorta like Object-Oriented Programming". They are all sorta right.
12:55:49<iand>tommorris: i use turtle when I introduce RDF: http://research.talis.com/2005/rdf-intro/
12:55:57<iand>btw, thanks for the camp
12:56:01<iand>it was a lot of fun
12:56:36<tommorris>i enjoyed it, but I'm taking some time off before organising another one. ;)
12:59:56<tommorris>http://tommorris.org/files/semanticcamp_rdfbeginners.pdf
13:00:14<tommorris>J:|SemanticCamp slides - RDF for Beginners (PDF)
13:00:57<tommorris>J: PDF slides from my RDF for beginners session at SemanticCamp. Done in traditional 'Keynote' style - ie. visual aid rather than long explanatory prose.
13:05:55<danieljohnlewis>yeah I think the SWEO project suggests N3
13:10:05<danbri>yep, semanticcamp was great :)
13:12:30<kwijibo>ACTION had a go at explaining RDF for programmers with RDF/JSON http://n2.talis.com/wiki/An_Introduction_to_RDF_with_RDF/JSON
13:12:36<danbri>tommorris, nice slides, one nit, "Directed acylic graph."
13:12:49<danbri>i'm no mathmo but i never understood acyclic in context of rdf
13:13:05<danbri>since we can have loops in the graph. is there some other meaning of acyclic i'm missing?
13:13:23<danieljohnlewis>lightning talks went quite well, we had: Jure Cuhalev (briefly demoing Zemanta), Oleg Lavrovsky (talking about Semantic Wikis and CMSs) and Libby Miller (talking about RDF and Widgets in Joost)
13:13:39<danbri>i have a lightning talk prepared
13:13:43<danbri>it goes like this
13:13:54<danbri>USE XMPP FOR SPARQL
13:13:55<danbri>USE XMPP FOR SPARQL
13:13:55<danbri>USE XMPP FOR SPARQL
13:13:55<danbri>USE XMPP FOR SPARQL
13:13:56<danbri>USE XMPP FOR SPARQL
13:13:59<danbri>END
13:14:17<danieljohnlewis>haha cool :-)
13:14:46<iand>danbri: you're always 10 years ahead of the rest of us
13:40:27<EtnaRosso>can i play with madness?
13:41:03<gromgull>EtnaRosso: sure: http://www.2flashgames.com/f/f-281.htm
13:56:51<yvesr>tommorris: yes, the danbri kissed himself example demonstrates such a loop :-) http://moustaki.org/resources/foaf.rdf
16:28:00<EtnaRosso>danja
16:28:11<EtnaRosso>hi
16:46:41<zerok>hi :)
17:00:00<mmealling>any WebDAV experts around?
17:37:58<tuukkah>bengee, should the autodiscovery link here work? http://semsol.org/semcamp/irc/logs/2008/02/18
17:39:36<bengee>hmm
17:41:25<bengee>the IRC logs are kept in a different store
17:42:20<tuukkah>i see - so maybe there isn't a simple way to view the triples
17:42:34<bengee>there is a sparql endpoint ;)
17:43:19<bengee>lemme see if I can tweak the link in the template
17:44:15<tuukkah>i don't have nice apps for viewing sparql endpoints =)
17:44:51<bengee>heh
19:26:41<bengee>tuukkah, gotta run, will have to do it tomorrow
20:26:38<DanC>does last.fm do openid?
20:30:03<DanC>setting up shop to do taxes on a new ubuntu box.
20:30:32<DanC>created some new key pairs while I'm at it... for family business. one for ssh and one for openpgp. sigh... why do they have to be separate keys?
20:30:55<DanC>hmm... ubuntu doesn't seem to have ssh server on by default; perhaps that's good...
20:31:37<DanC>WARNING: The following packages cannot be authenticated!
20:31:38<DanC> openssh-server
20:31:42<DanC>hmm... what's up with that?
20:33:34<DanC>ACTION gets the relevant clue in #ubuntu in a matter of seconds. :_
20:33:36<DanC>:)
20:35:26<DanC>connolly@gamebox:~$ ssh-keygen -l -f /etc/ssh/ssh_host_rsa_key.pub
20:35:26<DanC>2048 61:92:1d:6b:86:0d:41:ae:24:e5:1a:3e:57:ba:31:2f /etc/ssh/ssh_host_rsa_key.pub
20:36:16<yvesr>DanC: i don't think they support openid for now. but they did promise to support it a while ago. i don't know for when it is planned, though.
20:37:28<DanC>ok... so far I'm openid-only today... trying not to invest in passwords over the net
20:37:39<DanC>i.e. password-per-service
20:37:53<DanC>switching from del.icio.us to magnolia
20:38:00<DanC>or at least: trying out magnolia
20:38:29<DanC>http://ma.gnolia.com/people/DanC
20:38:53<DanC>K:|DanC's bookmarks on magnolia, an openid-happy social bookmarking service
20:39:23<dngor>Aww.
20:39:33<DanC>aww?
20:40:03<dngor>I work at del.icio.us.
20:40:20<DanC>do you have plans for openid?
20:41:06<dngor>I've been focusing on the "2.0" release. Let me check, though.
20:41:34<DanC>I'm a long-time delicious user; it would cost me a lot to leave altogether, so I'd be happy to stay if it grows openid suppport.
20:42:44<DanC>uh-oh... many tasty items on http://swig.xmlhack.com/ ... tough to stay focused on travel accounting and taxes with drupal/RDF stuff in sight...
20:43:30<dngor>Sometimes it's hard to debug things at work, for similar reasons.
20:43:38<DanC>has anybody showed adu how to add titles? let's clean up the in-your-face URLs, please
20:44:02<sbp>DanC: plus there's a new TWSW! http://blogs.talis.com/nodalities/this_weeks_semantic_web/
20:44:51<DanC>TWSW doesn't do it for me lately; I'm into story telling, and the bullets are either (a) stories I already know or (b) not enough of a story to engage me
20:45:35<DanC>I really liked the first few episodes, though
20:46:08<DanC>ACTION has a vague feeling that he asked sbp to read something when we last chatted...
20:46:41<sbp>hmm. I don't have the same vague feeling, but that's possibly just because I've been doing *so much* reading these past couple of weeks that it's got lost in the mire
20:47:10<DanC>H:see also Noah's recent draft finding on the self-describing web (my notes are in hardcopy; hope to transcribe soon)
20:47:29<DanC>must not have been that important then, sbp
20:47:59<sbp>note that @profile is a SHOULD in http://www.w3.org/MarkUp/Drafts/ED-rdfa-syntax-20080217/Overview.html#docconf now by the way
20:48:30<DanC>how about in the last call draft?
20:48:38<sbp>heh, oh dear. in the prose they say ‘There SHOULD be a @profile attribute on the head element that includes the value http://www.w3.org/1999/xhtml/vocab.’ and then in the example they do ‘<head profile="http://www.w3.org/ns/rdfa">’
20:49:00<sbp>there's a last call draft?
20:49:14<DanC>confucious say: the path to enlightment includes syncing examples in your spec with your test suite. :)
20:49:28<DanC>A:
20:49:48<DanC>that's where I got the idea there was a last call draft
20:50:29<DanC>A:| RDFa in XHTML: Syntax and Processing W3C Editor's Draft 17 February 2008
20:50:31<sbp>oh, it does say Last Call; missed that. same document as A: anyway
20:50:48<DanC>A: seems to be a last-call-to-be draft
20:50:53<sbp>phenny: tell Arnia http://salt.semanticauthoring.org/
20:50:55<phenny>sbp: I'll pass that on when arnia is around.
20:51:14<dngor>There's no open-id mention on our wiki. I'm adding it with a note that I've observed a long-time user trying other sites because of it.
20:51:32<DanC>thanks!
20:51:44<dajobe>omg salt
20:52:49<DanC>ACTION takes that as a prompt to take a look; misses the "omg" factor
20:53:24<dngor>What's your del. user name?
20:53:28<DanC>connolly
20:54:35<DanC>K:see also [http://del.icio.us/connolly|connolly on del.icio.us]
20:56:40<DanC>ACTION tries to focus... http://dm93.org/z2001/FractalAccounting
20:56:58<DanC>"Americans spend roughly 5.4 billion man-hours computing their taxes.(15) This is more man-hours than used to build every car, van, truck, and airplane manufactured in the United States."
23:48:50<daYZman>hi
23:49:40<daYZman>i found this title and it makes me ponder: Guided Analyses of Logical Inconsistencies Leads to Evolved Ontologies.. it is about a set of analytical techniques for ontology evolution
23:49:53<daYZman>should it be "Leading to" rather than "Leads to"?
23:59:18<TedThibodeauJr>no, I think the title's worded correctly ...
23:59:39<TedThibodeauJr>it encompasses completed, ongoing, and future efforts

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