Semantic Web Interest Group IRC Scratchpad

Welcome to the Semantic Web Interest Group scratchpad generated automatically from discussions on IRC at Freenode channel #swig 2001-2018 approx by the chump bot.

Nearby: IRC logs | semantic-web list | W3C Wiki (Recent changes) | delicious swigbot

last updated at 2007-11-01 21:22
 
sbp: Should now be linked in to the Exclusive Breadcrumbs Whitelist.
timbl: Aim of whitelist of only top exclusive spammers, otherwise to be inclusive!
 
danbri: Checks that it works OK much appreciated. Am hoping people will mostly use OpenID for access, but haven't mandated that. No anon-edits though (except you could use free public zero-auth openids of course).
timbl: Well done, Danbri, getting it back up, with spam protection
timbl: Well, re anon openid accounts, how about foaf+openid?
 
 
sbp: Yeah another one, but still it's not every day that one of these is released is it? This one is good because a) it's very anal about being specification compliant, and b) it uses Streaming Magics™ on top of expat to ensure speed and low memory usage. It's not as fast as rapper, of course, but it's faster than rdflib.
sbp: I figure I'll probably use this on the Semantic Web Survey that I'm doing. Previously I'd been using rapper in a subprocess.Popen call, but I didn't bother to set up error handling and streaming and so on for that—it's much nicer to have a native python solution that's well tested and efficient.
sbp: Quick link to the module itself: rdfxml.py.
 
sbp: As far as I can tell, it's impossible to parse N-Triples because of an ambiguity in the grammar; and even if the ambiguity were resolved, that would mean that N-Triples isn't able to express all possible RDF Graphs!
sbp: Which, if this is true, is rather devastating for N-Triples.
sbp: The most obvious way to fix this, based on what rapper does when you attempt to coax it into producing ambiguous N-Triples, is to specify that > must be escaped as \u003E in absoluteURI.
 
danja: Open access to 447,036 e-prints in Physics, Mathematics, Computer Science, Quantitative Biology and Statistics
danja: a candidate for Linked Data?
 
karlUshi: Review of the last book of David Weinberger - Everything is Miscellaneous
karlUshi: About classification, information, and taxonomy
karlUshi: "Weinberger has not done his homework and his book shows it. He repeatedly starts with bad assumptions, so the book crumbles into a pile of anecdote and opinion."
dajobe: and bonus dismisal of Shirky
 
Created by the Daily Chump bot. Hosted by PlanetRDF.