I know NASA used Reed-Solomon codes for the old Voyager probes. Maybe they're using something more efficient these days, but I'd have to imagine they'd be using error-correcting codes of some sort in whatever custom protocol they've devised. It would be ludicrous to use simple error-detection (necessitating a retransmit) at that latency.

As for the software itself, the Mars rovers just run VxWorks, right? Once you've got the code uploaded I'd think it'd be as simple* as restarting the process.

* Yes, I'm sitting in my pyjamas eating a bagel right now and describing what NASA does as "simple". Suck it!

That's IU. IO looks like an E with umlauts.

We speak English, not Latin. The Latin word "virus" has no plural. The English word "virus" does have a plural form, which is "viruses".

That's not entirely true, depending on what you mean by "independent". So long as you are a member of SOCAN and have music tracked by SoundScan, you're eligible for the levies, regardless of whether you're signed onto a major label. This flow chart (warning: PDF) describes the pay-out structure.

The media have been kind of lacking here, though. I have no idea how this pay-out scheme works in practice :(. Go go go investigative journalism!

No I think they got this going the right away around. The Internet will finally reach its full potential once everyone can filter out everything except .xxx domains.

No one has billions of dollars in chequing. These people have billions of dollars in investments. The mainstream media likes to talk about these investments in terms of dollars because it makes for cute headlines, but for the people involved I suspect it has more to do with control. Controlling billions of dollars worth of organizations means, for better or worse, you can help shape a sizeable chunk of the world.

Not true at all. The disparity between consuming resources to attack and consuming resources to defend is the crux of a Denial of Service attack, which is essentially what these drones are. If you know that you can force your enemy to spend $6000 for every $1 you spend, it's pretty easy to drag them into attrition.

I don't see why the summary is misleading or why a desktop machine would be the only measuring stick worth considering, especially when you think of how seldomly Linux is run on the desktop.

This is like saying, "My toaster runs linux and it can boot instantly!"

What would be wrong with that?

They poured a tonne of work into making this happen. Just because they control the hardware their hard work isn't worth anything? I think it's pretty cool what they've been able to do, someone no one else in the history of Linux has ever been able to do.

That's a really good point! You could have an acquaintance (you don't really like the guy, but hang out with him, know his living situation). Him posting to the world that he's going to be out of town for the next couple days could be tipping point.

The site doesn't tell you whether everyone in the household is gone, only if one person in the household is gone. A robber would still have to peek in the windows and do whatever it is robbers do to make sure the house is empty. But they could do that just by walking around some random neighbourhood and peeking in random windows; they don't need Twitter to tell them to peek into someone's windows.

It was a comment directly in response to "the iPad is the netbook of the future". Are you saying netbooks shouldn't be computers?

If the different colours are perfectly parallel, then there is zero movement in the upper layers and they only look parallel due to how the data is presented (stacked). In order for them to be "synchronized" you'd have to see the layers diverging from one another, not parallel to one another. You can a little bit of this, but not much. For instance, between December 2006 and March 2007, Office sales diverge a wee bit from the layers underneath. The Servers and Tools seems to stay completely flat, maybe even shrinking a bit.

Really it's just Windows sales going up and down and the two layers on top of them not doing very much.

I have some really stupid cheater stories, of course. Like the guy who handed in verbatim his buddy's assignment, including his buddy's name and student number at the top of the comments, and still kept denying it. Those are the easy cases to deal with and, while kind of funny, not very meaningful. If they fess up and apologize and co-operate you give them negative 100%; if they don't you give them an academic offence (which means automatically failing the course and automatically getting expelled from the university if they get a second offence).

There's definitely a lot more cheating going on than I can find evidence for, which I've come to accept is mostly out of my control. My only two real strategies for combatting it are to make the assignments as exciting/interesting as I can so that students don't mind doing them so much, and mandate that you need at least 50% on the exams to pass the course so that you can't get through just on the assignments.

The one that still affects me is a girl I had in a first-year course in my first year of TAing. She was one of the "cool kids" (a little slow transitioning out of her popular high school kid days, you know). Her friends sort of lived for partying and never took any of their courses seriously: most of them ended up failing the CS course. It was a lab scenario where it was scheduled for 2 hours, but if you finished your work in an hour (which many did), you left early. After all her friends had finished the bare minimum and had left, she would stick around, though, for an extra half hour or hour, asking questions and redoing the work she'd already done to make sure she understood it at all properly. She wasn't doing insanely well--I think she was carrying a mark in the course somewhere around the high 70s--but definitely better than her friends and she was putting a lot of work into what obviously didn't come naturally for her.

Then came her final assignment of the course. It was gorgeous! It was their only assignment where they had to incorporate object-oriented design (the course was taught in C++) and she did everything perfectly. It was definitely the best assignment out of the entire class, and we had a lot of really smart cookies in the class. I wrote glowing praises all over it. I ran into her at a bus stop a month or so later and gushed about how impressive her assignment was and how much work she must have put it into it and she just quietly smiled and didn't say much beyond "thank you".

Then about a year later I ran into one of her friends and the topic of her came up again. I said again how impressive it was and said "either she found someone else to write it or she put a lot of work into that". He got awkward and said "I don't want to get her in trouble, but let's just say she didn't put a lot of work into it".

It really did a number on me. I was at the end of my Master's degree at the time and it through me for a loop, wondering if I even wanted to come back for my Ph.D. Research is okay and all, but really the only reason I was in grad school was for the teaching. I eventually did go back and things have gone well, but it surprised me what effect one (previously) good student cheating can have on you :\

Yup, it's changed in Finder. Find a document, either right-click or go to the File menu, select Get Info, change the application in "Open with" and then select "Change All".

Indeed. One can't be totally certain (calculating entropy exactly is undecidable in general), but the entropy rate of English is approximately 1 bit per letter (give or take half a bit, so says Claude Shannon).

Assuming your passphrase is English, it would have to be somewhere around 30 words long to give 128 bits of security? That's essentially the entire first paragraph of this comment, a pretty long passphrase!