June 23, 2011

Slashdot Effect

When I woke up this morning, there were three surprising emails in my inbox, informing me about three comments someone anonymous had left on my blog. The surprising part was not just the three comments, but the fact that they actually seemed helpful.

The answer was in my list of http referrals. They were all from hardware.slashdot.org. Turns out there was this article about inscrutable logos on the back of devices, and someone linked to my article in the comments.

So well maybe that was not the real slashdot effect, but it was great to have a traffic spike that was 10% of my total traffic in a single day.

Apparently the slashdot effect follows a double peak - the first when the link was posted, and the second at around 8am EST when east coast comes in to work.

The effect seems to be petering out now, and the visit rates are falling back to my regular. But it is leaving behind a graph where the scale is stretched so bad that the usual seems insignificant. Maybe the real effect of the slashdot effect is not on the server, but the fact that it leaves you wanting for more.

June 17, 2011

Bioluminescence: The language of light

One, if not the biggest highlight of our recent trip to Puerto Rico, was the night kayaking in the Fajardo Bay.

Cameras were not allowed on the trip, but even if they were, I am not sure I could do justice to the unbelievable feeling of kayaking through water that lit up like a glow stick on contact. The cool blue color was not only breath taking, but it evoked a feeling of true astonishment. This giddy feeling of awe and wonder is something that Edit Widder has experienced more than her fair share of.

Oh, and the best part about our kayaking trip? No need for a submersible.

June 06, 2011

Cell phones & Cancer (infographic)

It was a strange feeling when I heard about CNN's breaking news that cell phones caused cancer. I had long known the need for caution, when it came to microwave radiation. But singling out the low power of a mobile phone in this age of electro-magnetic saturation seemed too fantastic.

Yet, the first few paragraphs in the CNN article seemed to portray it as a done deal. As if the most damning of evidence had quietly piled up. The strangest part was that there was no new studies being cited.

It took a little searching around and spending some time with comments on sites, that are not CNN, for things to start making sense. There were no new studies. Instead, based on the existing data (and wishy-washy footnotes that had been known for years) WHO had decided to formally include cell phone emitted EM radiation in the "needs confirmation" category. Yes, lead is in that category, but that is lead. Not inorganic lead compounds. Those are in another category which includes those items that actually do cause cancer.

The CNN article made it seem like cell phone radiation belonged to that other list.

Nevertheless, at the suggestion of a colleague, I figured the best way to reasearch this would be to put an infographic together. Here it is, on the left. It quickly introduces the EM spectrum and the ionizing vs non-ionizing nature of different radiation. And how the amount of energy emitted by a cell phone is limited. Before citing the tremendous body of research that has found no clear indication of a linkage between mobile phones and cancer. And finally a quick review of the grouping system used by the WHO and why Group 2B does not mean the sky is falling down.

This is one of the longest infographic I have made. Click for the full size.

June 03, 2011

Liberating data using Scraper Wiki

Of all the wiki sites that sprung up after the original, one of the most useful and positively cool is ScraperWiki. Scraper wiki is an attempt to liberate data from websites and pdfs and instead populate spreadsheets with them.

There is a lot of data available on the net. But its value is severely limited by the fact that you cannot do much more than just browsing it. When you move data from a html page or a pdf file into a spreadsheet, suddenly the value of the data goes up many fold. Now you can analyze the data, sort it, look for trends and coax information out of it. ScraperWiki aids in the first step by scraping web pages and moving data into usable data sets.

ScraperWiki is two things. First it is a web-based compiler and reusable libraries (in Python, Ruby or PHP) that allows you to write and run a scraper. Second, it is a wiki store of scrapers written by others that you can then update, reuse or just run to get data.

There are quite a few interesting scrapers. This scraper collects data from weather stations across all of Germany, while this one collects the Location IDs from Weather.com URLs. Weather is not all scrapers do, this one for example collects basic info about all MLB players, while this one is an massive database of all soccer WorldCup matches.

Of all the untold millions spent by governments and corporations on digitizing their data and making web pages, a decent portion that went towards making html tables out of data sets. ScraperWiki is an attempt to reverse that. Cheers to liberating data from the shackles of the web.

June 02, 2011

LOOM (Short Film)

Like short stories, short films are quick, easily consumable, bite-size experiences. But, lately, it seems as if the short films are going toe-to-toe with their longer cousins in terms of production quality and subject matter depth. LOOM is a brilliant example. As an animated short, this is a gut-wrenching yet life-affirming snippet of the world around us. Watch it below:

Loom from Polynoid on Vimeo.

Polynoid, the company that produced this has a few others they created before. Including their violently fascinating and creepy snail movie and the weirdly funny flap flap.

June 01, 2011

+1 and the persistent Like

+1 is Google's latest attempt at cracking social. After the famously obscure Orkut, and the disaster that was Buzz, it was about time Google got it right.

With +1 Google seems to have gone about social differently. +1 is a highly scaled down version of a social network - the opposite of Facebook. Facebook built the interaction feature-set first, then used the Like button to spread it. Google's approach seems to be focused on building out the +1 button and eventually coalesce the rest of its sharing services around it.

+1 has two things going for it - it is persistent and contextual. Persistent because unlike the Like button, the core idea for +1 is not to broadcast the action to everyone. When you "like" something, that act itself is shared by Facebook. Which works for Facebook, because communication is what Facebook is all about. But +1 is more persistent; it hides in the sandwich layer between web-content and you - the search engine.

Persistence is important, because this shifts the playing field away from conversations - which Facebook and Twitter are good at, to algorithmically mining history - something Google is great at. This is where context comes in. Google owns your landing page on the web: the search results. It is a powerful page, and is also contextual. Unlike the static Facebook, Google's use of +1 can morph itself to add context to what you are in the mood for at that time. Unlike a cacophony of likes, you instead get the few +1's that are highly relevant to what you are doing at that time.

This is the strategy that worked for Google in ads, and the bet is that it will work for social as well.

The problem for Google's social has not been building out social feature sets. The biggest impediment has been changing the nature of social to fit with Google's strengths. +1 could well be the game changing strategy that Google so desperately needs.