3/20/07

"Papers" on Mac OS X

See this "Paper" for Mac OS X, that let you organize and read PDFs. It looks pretty useful and I am very tempted to try it.


Basically, there are two kinds of approach to organize a large collection of files:
1. file system approach. It works for everybody by copying the metaphor of a tree structure. This approach is good if you browse more than you do search.
2. List structure, a.k.a the iTunes approach. This is great, if you search more than browse.



I like iTunes, and its search feature is a godsend>. If I want to listen to Beethoven, I just need to type "Beethoven" and all his 1 to 9 symphonies are all there. I would not want to go to the file system to drag beethoven music into Winamp.

Obviously, Paper is looking to do an iTunes approach. The pre-release app looks beautiful, but it runs very slow after importing 50 PDFs. Probably memory alloc/retains bugs? Yea I am sure they will fix it in a few weeks, but...

... I am still stuck in the "file system" approach, as I tend to put PDF books of different publisher in the same folder. And Spotlight serves me well for now? Everybody has resistance switching to something new afterall.

3/18/07

Gapminder at Google

TED 2007 has Hans Rosling's passionate speech about opening useful data to everyone.

Now you can see the animated stats interface on Google.

References:

  1. Gapminder World 2007, at Google

3/7/07

Jeff Hawkin's Numenta opens its "Numenta Platform for Intelligent Computing"

Just woke up and read Slashdot that Jeff Hawkins' Cortex Sim Platform is now available.

I did read Jeff's On Intelligence book about his idea of HTM, though I forget pretty much everything by now :-). But I am pretty excited about this, especially the 3 components have python bindings.

Go check it out.

References


  1. Numenta Platform for Intelligent Computing (NuPIC) downloads

  2. Introduction to Numenta's Research Release from Jeff Hawkins

  3. Numenta Blog