According to the summary, the GPU enhanced version uses ECB:

"A demo in its current source repository is a modified eCryptfs, which is an encrypted filesystem used by Ubuntu and other distributions . . . .However, both the GPU cipher-based eCryptfs and the CPU cipher-based one are changed to use ECB cipher mode for parallelism. "

I hope this is just a proof-of-concept design because ECB mode should not be used for this purpose. Wikipedia provides a pretty obvious example of the weakness of ECB mode:

"The disadvantage of this method is that identical plaintext blocks are encrypted into identical ciphertext blocks; thus, it does not hide data patterns well. In some senses, it doesn't provide serious message confidentiality, and it is not recommended for use in cryptographic protocols at all. A striking example of the degree to which ECB can leave plaintext data patterns in the ciphertext is shown below; a pixel-map version of the image on the left was encrypted with ECB mode to create the center image, versus a non-ECB mode for the right image."

http://en.wikipedia.org/wiki/Block_cipher_modes_of_operation#Initialization_vector_.28IV.29

No, 1 of 794 people identified were ultimately arrested of SOME crime, but not necessarily convicted. The article states that 40% of the arrests were immigration related, and a great deal more were drug related. Thus, the real number of false positives is actually much higher if you are only including legitimate threats to security.

Just to be clear, the paragraphs starting with "Members of Congress" and "Experts agree" are direct quotations from CNN. The remainder of the post contains my opinion. Sorry for any confusion.

According to CNN, the TSA is actually more ineffective than I initially thought:

False Positives-

Members of Congress also expressed concern about the number of "false positives" -- people flagged for additional screening that resulted in nothing being found. For every person correctly identified as a "high risk" traveler by (the behavior detection officers), 86 were misidentified, Willis said. At random screening, for every person correctly identified, 794 were misidentified.

Effectiveness at detecting terrorists-

Experts agree that the fact that there is an extremely small number of terrorists makes it hard to evaluate the effectiveness of behavioral observation programs. The Accountability Office said it looked at 23 occasions in which 16 individuals -- people later charged with terrorism-related activities -- passed through high-threat airports. None is known to have been identified. But it is not known if the behavior detection officers were working at the time, the agency said.

So, in the best case scenario, for every person ultimately charged with a crime (not necessarily convicted) 86 are misidentified. And that is using "trained" behavioral analysts. Most TSA searches are random, which results in one charge for every 794 false positives. Note also that nearly 40% of the charges are immigration related. Most of the rest are probably drug related.

The TSA can't point to a single incident where its random searches or behavioral analysis actually has prevented a terrorist attack. Despite their utter failure, the TSA plans to spend another $1.2 billion over the course of five years on behavior analysis techniques.

Source: http://www.cnn.com/2011/TRAVEL/04/15/tsa.screeners.complain/index.html?hpt=C1

I can only get 40 Mbit (5 MB/sec) on my 802.11n wireless router. However, I routinely get my rated speed of 50 Mbit/sec (6.2 MB/sec) down from my DOCSIS 3.0 connection, which peaks of 8 MB/sec. Of course, such connections are a rarity. Nonetheless, Gigabit ethernet is cheap and is far more reliable than wireless. Wireless connections just can't maintain multiple, high-throughput connections while wired ethernet can.

Also, I can easily get local transfers over my gigabit LAN that surpass 100 MB/sec. Unlike 802.11n, Gigabit doesn't have 50%+ overhead. I don't see wireless competing with wired ethernet anytime soon. Anyway, by the time Gigabit wireless is available, 10 Gbit Ethernet will be mainstream.

I'm currently signed up for Comcast's new 50/10 Docsis 3.0 connection. I get a constant 50 Mbit on Bittorrent with bursts up to 70 Mbit. Upload bursts to 20 Mbit and provides reliable 10 Mbit speeds. Not bad for cable. I am told that there are no data caps in my area but I won't know for sure for a few months :P.

However, FiOS still has better upload speeds and no download caps. Verizon offers 35/35 (around 43/35 in practice) in most of its service areas for $100/mo. or $115 with its ultimate HD package.

Was it a Plasma or LCD based screen? I was able to demo two 3D televisions at Best Buy. One was horrid (the LCD) while the Plasma had a very pleasing 3D effect without the traditional ghosting and blurring that you normally see at the theatre. I guess it depends on the implementation.

And if it was incorrectly calibrated by the Fry staff, what chance is their that Joe Consumer will figure it out. There are still people that think they're watching HD programming because the logo on their 4:3 27" 480i TV says "HDTV".

Odd, it scrolls fine for me in Chrome 5.0.375. There is a slight delay while the page fully loads, but after that it scrolls smoothly. Firefox is smooth immediately, but the delay is at most 1/10th of a second.

What build of Chrome and OS are you using?

I'm on Windows 7 and am using Chrome 5.0.375. I compared against Firefox 3.6.3.

I'm using Adblock, FlashBlock, and Session Manager in Chrome. While Adblock isn't as good as Adblock plus, it works fairly well. I certainly wouldn't say Chrome lacks a base of plugins.