Monday, October 25, 2004
Sunday, October 24, 2004
Its not a terribly important question, but I think it is an interesting one.
Monday, October 11, 2004
Back in high school, a physics teacher of ours said that hypothetically if one wanted to, one could crush a beer can using an induced magnetic force within a coil of wire. Just as a hypothetical example, of course. The required energy would be way too high to realistically do something like that.
Of course, high school students have selective hearing. Andy and I (eventually mostly andy) plotted on how to accomplish this feat. The trick was storing and releasing enough energy in one quick shot. To this end, one needs a capacitor. A big one. Our initial plan was to get a few hundred flash bulb capacitors from disposable cameras. These were free for the taking at local drug stores and other such places. Took alot of work, but many shocks later, we had a capacitor array. Only it didn't work well enough.
Ultimately, Andy continued the quest - buying a capacitor and acquiring the transformer required to step up the voltage. Yesterday, he showed me the fruit of the hard work. A can crusher for the not-so-masses. It takes about 2 minutes to charge up, and requires earplugs for the firer. Nevertheless, the resulting can is pretty cool looking:
Also, a close up shot.
Thursday, October 07, 2004
http://images.google.com/images?q=site%3Awww.poked.net+russia
From the count, it shows that google is missing only about 2 photos out of 579.
So there it is folks: my personal googlified photo gallery
Wednesday, October 06, 2004
So, the story goes - Yahoo created this individualized search engine thing that blocks websites. I decided to show off and do a one up on them with a google feed - over the course of one evening.
Well, I actually had this project in the back of my mind for a LONG time. And by back, I mean way back. Back as in hidden behind the spiderwebs in the shed in the far corner of your yard. That yard which happens to be in your summer house in sweden. That far back. But it has been there for awhile, and the Yahoo thing reminded me, so I all of the sudden decided to create it. The Signal-To-Noise metric is a pretty new addition to my concept of the project though. I had in mind something even more simplistic - but the SNR works exceptionally well.
If you don't know what I'm talking about, go to the site above.
My next thought is not that you should classify samples yourself but rather the system should get more results (say 100) and make refinement suggestions by self-organizing the pages into similar groups and finding refining words for each of those groups. I think I need to move away from PHP to do this though as it really is asking for a multi-threaded approach to perform the Google queries fast enough.
Ok, enough techno-rambling. I think it is a pretty good project for one evenings work.