In that XKCD he doesn't treat characters independently. Instead, he assumes that each word provides 11 bits of entropy (i.e. assuming uniform draws from ~2000 words), giving a total of 44 bits. That's far less than the (26^20) you'd get if you treated the characters as independent random samples.

It's Hitachi! Can't they just wave their Magic Wand and make the nuclear waste go away? Think of the buzz that would create!

First, I construct medium-size piles via insertion sort. That works great as long as there are few enough things that I can spread them all out and see where to insert new ones. Once that gets crowded, I stack that into a pile and put it aside. Repeat until every document is in a sorted stack.

Then I merge-sort the stacks.

All in all, I find it a reasonably efficient method.

It's not just what RAM usage is today. It's also what RAM usage *will be* in 2 or 3 years. I certainly want my laptop to still be useful after only a couple of years!

It has a nicer screen than a Macbook Air, at least.

There's a higher-CPU higher-disk version for a couple hundred more. I agree about the lack of RAM, though.

I think this is excellent. All the laws should be enforced, and if people don't like the results, they should change the laws.

Unenforced laws make everyone a criminal; then the law can be used as a weapon against anyone at any time, giving the government too much power.

I think for my purposes, "Use Julia" will eventually be the answer. But I'd be happier if Python could just do what I want.

I take your point; Python isn't the tool for proper threading.

Nevertheless, I think your claim that "people who need threads are already using other technologies" isn't true. I think people keep butting up against that need as their projects grow, and it forces them to (painfully) move away from Python. I think Python could better serve those users with good parallelism.

And Jython doesn't work with numpy, which is one of Python's best features. The JVM is great for parallelism, but numerical computing on it isn't very fast.

The main thing that keeps Python from being really useful for my projects is the Global Interpreter Lock (GIL). I would love to write Python for my data-intensive code, but it is impossible to get really good parallelism with Python; the multiprocessing library isn't a magic fix because then I have to move all my data back and forth between processes.

When, if ever, should I expect to be able to use Python to do parallel data processing? What is the priority for this, and what would need to be done to make thread-level parallelism possible?

Sorry, *Fjandr* specified tip.

I think you're right that paying in cash is usually the best option, if for nothing else than relieving the business of the credit card fee. But I also think that "tipping in cash" implies tipping in cash on top of a credit card payment, usually for tax avoidance, and that's what bothers me.

Those are many good reasons to *pay* in cash. You specified *tip*, and the only reason specific to tipping that I can think of is tax avoidance.

It's better only for the server, and it's only better because it helps them avoid taxes. In fact, that's worse for everyone in the country who is not the server.

I really like the Koss Plug series (e.g. http://www.amazon.com/Koss-Plug-In-Ear-Headphones-Black/dp/B00001P4XA). I'm not sure they meet your stringent audio requirements, but I think they sound fine, they're cheap, they fit in your ear comfortably, and they provide a good amount of sound isolation--enough that I feel I can safely listen when riding the train.

Their biggest issue is that because they're really *in* your ear, you can hear when the cord bangs against things. I don't mind, but you may. But for $12, you can pick up a pair and decide what you think.