Channel #semsol: Logs

This is a public chat log generated from the #semsol IRC channel.

10:46:21 bengee: bengee eventually finds a not-yet-taken name for his rdf browser and editor
10:47:35 danbri: beenbringpain?
10:47:41 danbri: er, beerbringpain
10:47:47 danbri: i bought that this morning, sorry
10:47:54 bengee: heh, cool
10:48:58 bengee: it'll be OS and go under .semsol.org, so there is no immediate need for domain
10:49:16 bengee: but grebr.com was still available
10:49:39 bengee: *Gr*aph *E*ditor and *Br*owser ;)
10:50:23 danbri: a nod towards http://en.wikipedia.org/wiki/G%C3%B6del,_Escher,_Bach ?
10:51:37 bengee: heh, no I'm not that smart
10:52:13 bengee: but "grab" and "grep" are nearby
10:57:33 bengee: bengee tweaks ARC's mf extractor to accept @alt in img tags as vcard:fn values
10:58:02 bengee: proper name extraction from twitter at last
11:14:36 danja: http://www.urbandictionary.com/define.php?term=greb
11:15:17 danja: (not reliable - like domain names, anything with <5 letters is already taken for some made-up slang meaning)
11:16:47 danja:
11:17:50 bengee: heh
11:17:52 danja: hmm, it's in wikipedia, mostly for a 1980's thing
11:19:43 danja: [[ The term "Greebo" appears in Terry Pratchett's Discworld novels, as the name of a cat. Pratchett's definition of a greebo is "Someone who wants to be a Hells Angel, but doesn't have the style." ]]
11:23:36 bengee: then "grebr lets people w/o a clue about rdf tech become semweb weavers" ;)
11:27:00 bengee: my initial idea was "gbase - like freebase in a box", but then gbase has dozens of other uses already
11:33:43 bengee: damn, another bug in the turtle serializer: http://arc.semsol.org/community/irc/logs/2008/01/17#T11:16:47Z
11:33:53 bengee: leading, double-quots
11:34:14 bengee: bengee fixes
11:51:49 bengee: arc-bot, quit
11:53:24 bengee: "a test" from the 00's
11:53:30 bengee: 'k
17:49:09 peterkz: Hi all! Is there an easy way to have the ARC store updated from a directory of files (similar to Virtuoso)? I am thinking about something monitoring the file store for changes and re-loading changed files into the 3ple-store.
17:50:07 bengee: hmm..
17:50:53 bengee: sounds like an ARC plugin use case.. ;)
17:51:05 peterkz: I guess this is how many would like to feed their store with data?
17:51:14 peterkz: Ah, yes.:-)
17:52:27 bengee: we've been chatting about background processes and crawling yesterday on #foaf
17:52:42 peterkz: So, a bit of configuration (root filepath), maybe some filename pattern matching (like *.xht2) and the context URI could be the filepath (e.g. LOAD ... INTO </my/file/path/file.xht2>)
17:53:36 bengee: you'd just need a long-running php process that monitors the given directory, and does the DELETE + LOAD on changes
17:54:08 peterkz: Yes, it sounds simple. I am not sure what features there are in PHP to easily monitor file system changes. Or do I just have to loop over a collection?
17:54:37 peterkz: ...of files.
17:55:08 bengee: I only know the loop approach
17:55:26 bengee: is the direcory http-accessible?
17:55:35 peterkz: Yes
17:56:21 bengee: some web servers offer a way to sort by last_modified in their directory view, you could use scraping to get the top x entries ;)
17:56:52 bengee: lemme check the php spec for a more solid way
17:57:36 peterkz: Ah, good idea. That should be a simple approach. Maybe there is some shell script to list changed files form the past x hours. That would be easy to call from PHP I guess.
17:57:55 bengee: yeah
17:57:59 danja: mortenf, what was the uri of your sparql result 2 table xslt again please?
17:59:09 peterkz: Here it was: find ~ -mmin -90
18:01:04 peterkz: If successful, this plugin could be a nice way to have a collection of files with RDF data that is in sync with the ARC store.
18:01:26 peterkz: Or maybe it doesn't have to be a plugin. It is just a script I guess...
18:02:26 bengee: yeah, could fit in a simple script
18:02:48 peterkz: I'll try it out and see how it works. Bye!
18:02:54 bengee: but could also still make sense as a plugin, so that you only have to include ARC, the plugin, and call go()
18:03:02 bengee: cool
18:03:53 peterkz: Yeah. Only difficulty is the long running PHP process. Mosts shared hosts will cap the running time I guess. But a cron job with a simple script could do the trick.
18:04:22 bengee: or self-reinvoking via http sockets
18:05:08 danja: bengee, do you have mortenf's table xslt URI? ...or what's the URI of these logs again? :-)
18:05:57 bengee: hmm, table xslt, not sure
18:06:12 bengee: logs are linked from topic
18:09:02 bengee: hmm, the logs' sparql endpoint doesn't find much with xslt in it
18:09:39 bengee: but I have the mail somewhere
18:10:10 bengee: ah, no, that wasn't by mortenf
18:13:24 bengee: ah, got it
18:14:21 bengee: danja, http://www.wasab.dk/morten/2005/05/sparql-html.xsl
18:14:54 bengee: via http://chatlogs.planetrdf.com/swig/2008-01-05.html#T18-51-54
18:15:35 danja: excellent, you're a life-saver bengee
18:15:48 bengee: welcome :)
18:36:42 bengee: oh, asemantics goes bbc