I am in one of my sleepy phases right now, and yet, my browser is stuffed once more with fascinating links to all kinds of cool things. So obviously, I have no choice but to do invoke the nuclear option, and…

RELEASE THE LINKSTORM!

Long Live Chaos Monkey

First up, there is this absolutely fascinating little article about an unusual Netflix app.

The app in question is a script called Chaos Monkey, and as the name implies, its sole purpose is to disrupt the Netlix Instant Service by shutting down servers and instances. They have essentially programmed their own saboteur and told him “Go. Destroy. Make us miserable!”

Sounds kind of insane, right? But it’s actually brilliant, and a marvelous example of exactly the sort of “next level” thinking that I adore.

The idea is that by loosing the Chaos Monkey on their systems, they can continually test just how well their system reacts to compensate for problems, and thus, Chaos Monkey teaches them to make absolutely everything as independent of other systems as possible, with redundancies on every axis and the dependencies down to an absolute minimum.

For too long, people have tried to get away with reaction based emergency planning in system design, where you make the simplest system, then wait for something to go wrong, then slap something kludgey together in a blind panic to fix that one thing, and then go back to not thinking about it till the next failure. And so forth, and so on, with plenty of people slapping corks into leaks but nobody thinking about how to build a better dyke.

By willingly and knowingly increasing the chaos and entropy in the system, the Chaos Monkey idea forces people to build systems that can handle anything and that prevents the dangerous “complacency and emergency” mindset from setting in.

Sure, work hard to prevent failure. But do not fall into the hubris of design which lures people into thinking their design is “foolproof” and therefore they can safely ignore the “What do we do if something goes wrong?” question.

Know that you are not perfect and therefore neither is your design. Like the article says, “assume failure”. And plan accordingly.

It’s just so sensible and intelligent, it makes me want to weep tears of joys.

Hey This Place Sucks Too!

Then there’s this little article about a concern for the far future, namely, would faster than light travel totally destroy the planet you’re traveling to?

You have to admit, that would suck.

And in at least one form of faster than light travel, the Alcubierre Drive, that is a distinct possibility.

When traveling via Alcubierre Drive, the spaceship would end up accumulating a huge buildup of high energy particles moving at the speed of light as it pushed through space and ran into random interstellar hydrogen and so forth.

Then the ship arrives and stops, and all those high energy particles are released from the warp field, and annihilate the very place you were traveling to in the first place. And the longer you have spent in warp, the more particles and the greater the damage. Yowch.

I can easily imagine a very tragic science fiction story where a group of space colonists keep traveling to planets that their research says should be teeming with life, but when they arrive, it’s just a lifeless rock of no value whatsoever. Then, they learn their Terrible Secret… it’s their very warp drive that is destroyed their potential colony sites! Oh, the guilt!

Of course, there is a perfectly obvious workaround. Just don’t come out of warp directly aimed at your destination. Aim yourself in some harmless direction, then turn your ship towards your destination. Problem solved.

A more intriguing idea would be the idea of doing this deliberately as a weapon. It would make a great “secret Kirk strategy” type plot-solving climax to a space opera type science fiction story.

“But what you don’t know, Professor Oblivion, is that I have had one of my ships cirling in warp for the last two weeks in anticipation of just this situation!”

Robot pathos strikes again

Finally, we have this little nugget of animation, entitled NO ROBOTS.

No Robots from YungHan Chang on Vimeo.

First, disclosure : my first impression of this short was a bad one, because the description Ebert gave it said something about “a future without robots”, which made me think it would be a high concept science fiction story about a society which gets rid of the robots on which they had become dependent, and what happens after.

And the first part of the animation totally supports this idea, so that when it devolved into a smaller story about a milk stealing robot, I was quite disappointed.

But on second viewing, I decided I quite liked it. I love the art style. It reminds me of some of my favorite pieces from Heavy Metal Magazine, from the always much better but little ballyhooed “non tits and gore” stories. A very European style, in a good way.

And I can’t stay mad at a story that ended in kittens. Happy kittens that now have a human protector to keep them and care for them and try to keep their robot saviour upright.

Granted, it’s a fairly obvious and totally schmaltzy ending. I expected the robot to be caring for a baby with all that milk, but kittens is an even better choice, because a human baby would have raised the specter of a human being raised by robots and would have made things kind of squicky.

So kittens it was, and that makes for a happy warm ending, and I have no problem with that. I am quite sentimental in my own way, and a sucker for legitimate warm happy feelings, and so I do not mind being manipulated a tiny bit to get there.

The world needs all the warm loving happiness that it can get. Anything that helps people connect to the wellspring of deep human compassion is fine by me.

Even if it takes a robot with some kittens.