Yoz [05.10.06 04:32 AM]

Here's a fantastic 3D Java visualisation of all the location-based subscription data collected from your mobile phones and sent to our geostationary weather satellites, which were all built and launched by Phil Torrone using Lego Mindstorms, five hundred abandoned X11 manuals and a Perl script. Phil will be presenting this in more depth at July's "WTF 2.0" conference - don't miss it!

  liza [05.10.06 06:43 AM]

My Ajax-driven folksonomy has rounded corners.

  Greg [05.10.06 01:07 PM]

This is probably one of those get a life moments for me, anyway I hacked up a quick script that looked at the last 6 months of archives, recorded word frequency, removed the 5000 most common words in the English language according to the first source Google gave me and then looked at leader and follower words. And the results of the British jury are (after a little human editing (I ignored words like O'Reilly, E-Mail, Blog)) ...

• Google (76 mentions) - 2nd most frequent interesting word behind only web (140), with Google Maps occuring together 21 times.
• 
  Maps in general, 34 times and an additional 22 for the word map as opposed to maps, although as said above 21 had Google as a lead word.
  
• Yahoo (29)
• Microsoft (24) - Surprised at this result and the Yahoo! one, I never got the impression from reading the blog it mentioned companies that much.
• Database (19), although this is probably down to the recent series.
  
• Flickr (18)
• AJAX (17) - well there is a shock ;-)
• Virtual (19) + Virtual Worlds, came in strong but has dropped off recently, but not as bad Perl has.




• Startup (12)
• Ruby (11) - and its strongest follower Rails of course (6)
• Maker (11) + Maker Faire (6)

It will be interesting to see how a brute force approach with just a little editing compares to your opinion.

n.b. when i talk about leaders/followers I'm looking at them after the common words have been removed. e.g. Ruby on Rails, gets 'on' removed and hence Rails is a follower of Ruby.

  Greg [05.10.06 01:23 PM]

Oh, I forgot the patch I thought of in the car home, so I have one last addition to my list. First of all the patch and those of you with regex-fu can guess whats coming.

[gem@buffy RadarTagTickles]$ svn diff
Index: rtt
===================================================================
--- rtt (revision 4)
+++ rtt (working copy)
@@ -38,7 +38,7 @@
 
 sub get_words {
   my $content = shift;
-  my @words = $content =~ m/([a-z'-.]+)/sig;
+  my @words = $content =~ m/([a-z'-.]+|2\.0 )/sig;
   @words = grep { !/^['-.]+$/sg } @words;
   @words = grep { length($_)>1 } @words;
   return @words;

And my addition is ...

• Web (140) - on its own fairly boring, but if we looked a the most popular follower '2.0', we see that 'Web 2.0' is mentioned 42 times.

  Nicolas Toper [05.10.06 03:05 PM]

folksonomy, tag, FlickR, social software... Actually why not a tagcloud on your Radar blogs. You would find out quite easily what are those buttons.

  Patrick Tufts [05.10.06 03:23 PM]

An O'Reilly Radar post titled "Web 2.0 creative commons alpha geeks' location-aware smart API RSS mashup" just might peg the meter.

  Xavier Cazin [05.11.06 06:19 AM]

Let's try Data-driven business: a human-centric pov ;-)

  Anjan Bacchu [05.11.06 04:41 PM]

hi there,

"inaccuracy" -- I wrote a comment to tim's response post regarding Pragdave's post the new way to publish books. I was not correct -- my comment did NOT get published.

So, you guys like people who're correct. It irks you when people quote you wrong or their quote is inaccurate.

BR,

~A