A backup tape was stolen from his car last month containing at least 770,000 social security numbers (with the corresponding names) for Ohio taxpayers. It also contained the social security numbers for another 64,000 state employees. Today the intern issued a statement with his side of the story.

Add OpenID support to Mailman accounts.
* This would allow one account to be used for all Mailman installations
* It would also get rid of the monthly email reminders Mailman often produces

Originally spoken on 11/9/2006, just posted online this week. After exploring a bit of these guys work for my undergraduate research I heard about this panel. Quite topical. It ranges from the early non-Luddite commune culture that the WELL grew out of, the making-to-test nerd environment that silicon valley embodies, through our current ire at the phone companies un-silicon-valley-like world (premonitions of The Steve's iPhone announcement), the Libertarian ethic, its ups and downs, and even a bit of discussion about Google's growing power and "what if the Government took it over".

The last reminded me of the note from Cringely this week that Google has leased huge quantities of fiber across the nation. He suggests they have plans to offer peering agreements to ISPs. This would make super-high-bandwidth internet usage by the masses sustainable for these companies who for years have been over-booking and over-selling their real capacity.

Inspired by NerdTV show with Dan Bricklin, there is now a website where visitors can upload data and assemble graphs from it to find interesting things.

Relatedly, I came up with a funny technique that I performed over a couple hours to generate a user growth curve of Kuro5hin.org with 35 (fairly solid) samples and posted the data. I took first posting date of two to five userid-sequential users until I was fairly certain that was the day that that region of users had joined. I didn't really expect the sampling to work but after I got a reasonable number of data points down I couldn't break it.

On a side note, I'm not sure why this particular graph (out of the twenty or so autogenerated ones), the most valuable on in the set, despite being rated well by me, is not on the front page for the dataset.