Yes, man-in-the-middle attacks still work. There are a number of other attacks that work also. This wasn't an attempt at explaining every single little detail of cryptography but rather answering the question which was asked- why we care about large primes for crypto purposes.

24 lines? Uh, that is a T1. T1 is available everywhere, although it might get rather expensive in some places.

In general, a T1 seems to be much, much less latency than any DSL I have ever seen. A lot fewer routers in the way. End result is that a 1.5Mb T1 is a lot closer to 3Mb DSL, maybe 6Mb in some situations. Having had a business on DSL a couple of times but mostly on T1 connections this has proven itself several times.

The real question on all of our minds though: "How far will it launch a pumpkin?"

The corp/alliance structure will really help with this I think. From the article, it sounds like as a merc you'll be hired personally. So yes, you might get hired to lead a random bunch of morons who won't do anything you say. OR you get hired along with the rest of your corp/alliance to beat up on the members of another corp.

In the second case, if the grunts you are commanding don't follow orders, kick them out the corp. Put black marks by their names. Tell everyone how much they suck. Refuse to command them again.

The question, I think, comes down to the DUST players. How many serious players will get involved in the corp/alliance structure? How many halo players will just want some laughs screwing everyone over? If there are enough of the former that you can avoid hiring the latter, this will be awesome. If the *only* way to populate your battlefield is to hire the casual players, then I agree, the RTS side of things will degenerate quickly.

But as long as there are some rewards for winning a fight, I imagine the serious players will gravitate towards the serious generals and form groups that are very hard for casual players to beat. All in all a win for user-driven content.

You're right that I did mix feature sets with interoperability. Both are valid points. There are still lots of things you can do with MS Office that you can't in Open Office. I'll be honest and say I use both. I like OO for basic stuff like simple word documents at home. I don't use it at work because the features simply aren't there. Impress and calc are toys compared to MS Office.

The only thing OO has going for it is the price and multi-OS support. It's quickly becoming slow and bloated though.

Eh, on the LHC at full steam, we have collisions at 40MHz. The ATLAS pixel detector, is an 80M pixel chunk (or rather ~28k chunks) of silicon. Admitedly, most of that data never makes it out of the on chip electronics, and it has to be triggered, and the pictures are VERY sparse (a few thousand pixels fireing in an event out of the 80M), but still. We can take those snapshots damn fast.

PONIEZ FOR EVERYONE!!!

Check out the sales of Eve online on march 10th. They are putting it out in a box set for the first time (well practically the first time). Before now it's been download only. If the number of people playing shoot up, that's a good indicator. Likewise if the box set falls flat.

Build a particle accelerator. Not a big one. I've seen homemade cyclotron on slashdot before: http://science.slashdot.org/article.pl?sid=02/10/20/1626204&tid=134&tid=14/ Point it at your cloud chamber when you're done, of course.

Superfluid He is also wickedly cool. If you can build something to house it, and pump on it until it gets cold enough, you should be able to do some cool experiments with it.

Though not as visually appealing as a cloud chamber, building a detector to measure the lifetime of a muon was one of my favorite undergrad experiments. Three scientilators stacked on top of each other wired into a bunch of electronics, along with the right formulas, and you can get a reasonable measurement. My prof gave us a Phys Rev paper describing how it was done years ago, access to the parts we needed, a scope, and a computer that had a labview application set up for counting experiments. We figured out the electronic logic from the paper, used the scope to debug and set all the triggers correctly, then had to figure out how to actually calcuate the lifetime from what we measured (along with systematic + statistical errrors). Hard, yes, but man we learned a good deal about real nuts and bolts experimental physics that quarter.