So prepaid works? They allow it? Also are there any issues with keeping my AT&T phone number with the prepaid plan?

You shouldn't have any problem with porting (unless you're under contract) or using a prepaid plan. "Even More Plus" plans are still available though no longer advertised and may save you $10/mo.

Also consider the T-Mobile reseller Simple Mobile. $60/mo gets you "unlimited" service.

I could do that. I was really hoping there would be something off the shelf I could use (even if I have to install it on a server.)

That would almost work but I'd need a way for the parents to view the spreadsheet without exposing it to the internet or requiring them to have gmail accounts.

So the answer is learning software!

It sucks he didn't put a date on that post. Kind of an egregious oversite, right.

Not sure if you're being sarcastic, but the date is right at the bottom of the post: "(Published Nov 10th, 2004)"

I've always wondered why we can't just blow up hydrogen bombs in a giant underground chamber and extract energy from the heat produced.

Guess what? That's exactly what the NIF plant in question is: http://wingolog.org/archives/2010/01/31/nukular

Much safer to just run a nuclear reactor. But I wonder... would a fission explosion-based reactor even be possible, let alone more efficient? I don't think so...

A nuclear reactor uses fission. My idea uses fusion triggered by fission (look up how hydrogen bombs work). Fusion is much more efficient that fission which is why my idea has merit.

Ok now that it's hooked up to a computer, how about using a genetic algorithm to find the perfect roasting settings? (As a bonus it could evolve over time to match your gradually changing tastes.)

As a home coffee roaster myself (including building my own roaster) I'm guessing that experimenting programatically with roast times would cause the smoke alarm to go off regularly.

I guess you could hope that later generations get better at this, but there's always the risk of a 5 alarm mutation ;-)

Hook up a smoke detector to turn off the roaster. If smoke doesn't clear in 1 minute, have it shoot off a fire extinguisher. Problem solved.

I just finished reading "Gray Hat Python". It was pretty interesting talking about fuzzers, and injection stuff. Is that what you guys do?

Very yes. (We're working on "Grey Hat Ruby", though I don't know if we're going to publish it on dead trees).

So dumb question. How do you make money doing that? Is it that software companies want their security tested so they hire you? (I wouldn't think most companies would be so proactive.)

Every fiscal quarter, tens of products ship (often after millions invested) for every one practitioner that can competantly write a fuzzer. That estimate may actually be conservative by an order of magnitude.

We work for vendors and enterprises (ie, normal companies). Vendors bring us in to beat stuff up before they ship it. Enterprises bring us in to reverse things and beat them up before they get deployed.

Yes, many companies aren't that proactive. But the industry is slowly getting dragged into security; key verticals like financial services and health care are starting to require documentation of penetration testing for anything that gets deployed, and Microsoft, Google, Mozilla, Apple, and IBM all have religion about software security process.

Do a blog post about this. I think it's a question everyone has had at some point. How much of the bezel is really neccessary? Can it be removed DIY?

Just get your buzz saw and slice em off. I'm sure it'll work fine, how sure I wouldn't want to say.

So you could make a thin stylus if you put a charge on it? I guess otherwise using a conductor it would have to be the size of your finger.

Might be a good idea for a third party product?