Archive for the ‘technofetish’ Category

Headlines accused of being interesting… and somewhat misleading

Wednesday, June 9th, 2010

The headline:

Google accused of criminal intent over StreetView data

Usually the BBC is pretty good about putting stuff like “criminal intent” in quotes in their headlines. Tho to be fair Google was accused of criminal intent, just not by anyone that matters. To wit:

Google is “almost certain” to face prosecution for collecting data from unsecured wi-fi networks, according to Privacy International (PI).

The search giant has been under scrutiny for collecting wi-fi data as part of its StreetView project.

Google has released an independent audit of the rogue code, which it has claimed was included in the StreetView software by mistake.

But PI is convinced the audit proves “criminal intent”.

When I real something like this, I tend to read between the lines and notice “Google accused of criminal intent by a private non-profit that badly needs some good publicity right now.

What Google will face is some police investigations.

“The idea that this was a work of a lone engineer doesn’t add up. This is complex code and it must have been given a budget and been overseen. Google has asserted that all its projects are rigorously checked,” said Mr Davies.

Actually, this is probably NOT the truth, Mr. Davies, as WiFi snooping tools and data loggers of open WiFi have been around a LONG time, since the advent of WiFi to be truthful. So writing a code that says “if encerpytion key exists don’t log, else log” for the WiFi data dumper wouldn’t be that hard.

Also, Google encourages engineers to work on their own projects. You’ve heard of their 20% time presumably, Mr. Davies, because you’ve studied the Goog so much.

For all we know, this WiFi dumping tool started out as internal auditing software, for them to drive around the Google campuses and find out if anyone was leaking data out of HQ. Cuz shit man, it sounds like that would be a good tool for anyone to have.

All I know is, back in the 90s, you could buy a keychain that would light up when a WiFi network was nearby. A keychain, in the 90s. Sniffing publicly available and broadcast through your body (talk about an invasion of fucking privacy!) WiFi data ain’t hard. Telling the difference between encrypted and non-encrypted WiFi ain’t hard either.

Oh, and this isn’t the first time they’ve been wrong about the Big G, so take their words with a grain of salt.

No, you hold on a minute, D. Pogue

Monday, June 7th, 2010

>I don’t know if you’ve seen the frantic blog headlines, but they boil down to this: Those big, greedy, monolithic cellphone companies have found yet another way to gouge us for more money.

Well, hold on a minute.

AT&T is a business. They don’t do stuff unless they think it’s gonna make them more money. End of story. They will give you a host of reasons but what it boils down to is the bottom line. They are a publicly traded company and they have a fiduciary duty to their shareholders to make money. Or the shareholders can sue them. That’s how capitalism works.

So, uh, yeah, they are trying to gouge you for more money. Not us, because I left them long ago due to their horrible (well, not so much horrible as non-existent) support for the technology their phones used (Oh, can’t plug that into your computer? Yeah, we don’t support that, sorry you spent the extra money for a phone that hooks up to your computer.)

T-Mobile caught me on the rebound and has served me well ever since. Long live Android and have fun paying the AT&T tax to use your shiny new iPhone 4. It almost multitasks but good thing it doesn’t really so you wont have to worry about it sneaking data down in the background.

UPDATE:

Well, I’m not too worried about the video-calling thing. You could videochat for 15 minutes every single day and still use up less than half of your DataPro allotment.

Oh, except you can’t because video chat doesn’t work on 3G, only WiFi. Currently anyway.

Another thing that grinds my gears on the whole thing is AT&T’s pricing scheme. $15 for 200 MB, $25 for 2GB. If you pass your initial 200, you pay $15 for another 200. That’s right, for 400MB you’ll pay what users are currently paying for unlimited. Sure, you can save yourself potentially $360 a year, but only if you stop using everything but the phone feature after you hit that 200, otherwise you’re just paying the same for crap. Currently, they claim, 98% of their users use less than 2 gigs a month. But if so few users are using ‘excessive’ bandwidth, why is their network so shitty and how will this pricing plan have those 2% of users paying for all that ‘excessive’ bandwidth that’s supposedly killing their network.

Why is their network so shitty that 2% of their users can kill it, and making those users pay an extra 5$ a month going to pay to upgrade their network?

It’s not. This is a play to extract extra money from those who are bad at math and just trying to save a couple bux up front, never realizing they’ll be paying 5$ more than the people getting 5x their bandwidth when they go over that measly 200 megs.

Drunk History of Tesla

Tuesday, May 11th, 2010

this is pretty funny, but look away near the end if you’re emetophobic.

We’ve talked about Tesla before. Yeah, people use his ideas to perform PURE AWESOME.

And here’s two-part Tesla coil harmony, appropriately playing Russian music.

R&Sie Showreel

Friday, April 30th, 2010

via papa ellis’ infosphincter

WHAT IS MEMRISTOR?

Wednesday, April 7th, 2010

THIS IS MEMRISTORRRRRRRRRRRRR!

Memristors are amazing. They can work with CMOS (your regular computer chips) and different scales, i.e. they can mix 100nm memristors with 45nm CMOS structures (intend to shrink to 5nm memristors with 45nm CMOS, as they are so awesome CMOS doesn’t really need to change any further to accommodate their awesomeness-imbuing powers), they can tolerate up to 20% failure rate. It’s damned near magic they way they work -- swapping oxygen atoms back and forth between a layer of TiO2 and TiO by changing voltage slighty, changing it’s resistive properties. The change is about .3nm thickness to either layer, depending on which way the voltage goes. And it will be much more efficient than regular transistor based computronics, you should be able to pull ten resistors out of a given design and replace them with one memristor. And as they’ll be able to pair 5nm memristors with 45nm CMOS, that means those ten resistors will actually hold enough space to put about 90 memristors. Holy crap, amirite? That’s a factor of 900X. Shittin your pants yet? This next decade is going to blow your mind.

Further reading here. Warning: deep nerd.