On the flip side, with every transaction being reported and no loop holes to exploit, the percentage taxed (0.3%) can cover the revenue otherwise lost. Still, privacy would remain a concern.

You'd be surprised. With vacations, overtime and the union all involved, it can get to be a real mess. The overtime especially as if it's not doled out exactly evenly/correctly, the whole schedule can be affected. Also, they simply have to hire one person for every 5 due to the 6-day schedule who works all the others' off days. They can cut a LOT of cost by eliminating this position altogether.

While they don't have to pay the staff anymore on Saturday than the rest of the week, there's sizable overhead in managing 5-day work weeks when mail is delivered 6-days a week. By going to a 5-day delivery schedule, they can save on all of the overhead and assign one carrier to one route all week. This makes things easier and more efficient all around.

Because conveniently, that (incredibly biased) site only shows data through 2007 after which presumably, they started losing money. In any case, the whole point of ending Saturday delivery is to remain self-sufficient.

At first glance this seems like a ridiculously good idea. What's the catch?

That stuff only comes on Wednesday because the backlog of mail from over the weekend has finally subsided by Wednesday. If they got rid of Wednesday delivery, you'd just get the same crap on Thursday.

See http://slashdot.org/story/10/03/31/1834255/New-Method-Could-Hide-Malware-In-PDFs-No-Further-Exploits-Needed

Caterpillar is one of our customers, yes. I work in a variety of industries, but mostly with heavy equipment [read: very large vehicles].

Care to back that up with any facts? Matlab/Simulink definitely DOES drop into such systems. In fact, by using RTW Embedded Coder as mentioned by another poster, you can (and usually do) generate the entire binary in Simulink.
See here: http://www.mathworks.com/support/solutions/en/data/1-4Z7I7F/?solution=1-4Z7I7F

And it conforms to a myriad of automotive and software standards that I'm sure involve more testing than the OP would have time to accomplish in any reasonable amount of time.
See here: http://www.mathworks.com/automotive/standards/maab.html

Or the first guy is using it wrong and taking the chance of introducing even MORE bugs (more cooks in the kitchen) while the second guy is relying on code that has been tested time and time again, not only by the Mathworks, but by all of their customers as well. Tell me, when writing code for Linux do you re-evaluate every line of the kernel or treat it as a black box? One of our largest customers (a Fortune 100 heavy equipment manufacturer) relies on generated code to control their engines. And these are big engines. The Mathworks produces very solid code allowing developers to create control systems very quickly that are time-tested to be reliable. That being said, that doesn't mean Toyota simply didn't connect the blocks wrong in this case. A human is still responsible for the logic.

Actually, it seems they index something at least. Take a look at this search: http://www.google.com/search?hl=en&client=firefox-a&rls=org.mozilla%3Aen-US%3Aofficial&hs=vsl&q=litigious+bastards+jones+day&btnG=Search

At the end of the article is a link to a follow-up by the assistant news director. It's not quite an apology, but it demonstrates that someone in the Ubuntu community got through to someone at the news station: http://addins.wkowtv.com/blogs/behindthenews/

I noticed the format change too, and frankly, I liked it. I think in a lot of previous interviews, the questions could have used some editing. But the best part about the new format in my mind, is that they asked followup questions a couple times -- something the old format didn't really allow for. I liked it. It was much more coherent.