- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Thu, 30 Jul 2009 16:01:20 -0500
On Thu, Jul 30, 2009 at 3:45 PM, Aryeh Gregor<Simetrical+w3c at gmail.com> wrote: > On Thu, Jul 30, 2009 at 4:34 PM, Jim O'Donnell<jim at eatyourgreens.org.uk> wrote: >> I think Google News Timeline is worth?mentioning?here as an application >> which already does this >> http://newstimeline.googlelabs.com/ >> It shows >> events?going?back?to?the?late?Middle?Ages.?I'm?not?sure?how?they've?harvested?the?dates?from?wikipedia.?Perhaps?by?using?microformatted?dates? > > Probably by looking at the category markup Wikipedia uses, which is > specific not only to MediaWiki but the particular decisions of the > English Wikipedia editing community. ?<time> certainly wouldn't help > here -- this application doesn't need a way to say "this text string > denotes a particular time", but rather "this event happened at this > particular time". ?You need to use Microdata or RDF or such for that > kind of relationship semantics. Nod - automated mashups of that kind are outside of the ability of <time>, even if extended to handle those dates. It seems that the useful things to do are: * directed mashups, where you can manually point an app at particular dates that you catch across the web and automatically slurp them. * more reliable temporal searching, at least for pages that include this kind of markup. Rather than just searching using "2000 BC" or whatever and hoping that the target page includes that phrase, the search engine could also take note of manually encoded times and use these to help filter the returned results ~TJ
Received on Thursday, 30 July 2009 14:01:20 UTC