Forgot your password?
typodupeerror

Comment: Re:And less than four years later... (Score 2) 204

by lennier (#47496473) Attached to: Apollo 11 Moon Landing Turns 45

For one, we will eventually exhaust all the resources on this planet, and our species will become extinct if we cannot - at the very least - successfully extract resources from other worlds. We really need to find a way to actually live on other worlds if we are to continue to exist.

Actually, it's fairly easily shown that if we continue our current exponential rate of population growth and resource usage, we'll use up the entire Milky Way Galaxy in 2,500 years. That's assuming nonexistent magitech FTL drives which contradict our current fundamental physics theories.

Or, we could stabilise our short-term rapid growth and learn to live on the one accessible habitable world we have, like we did for the past few million years. Our choice.

By the way, any future that has economically viable space colonies in it will also have economically viable greenhouse cities in Antarctica, the Sahara and the Australian outback first. Because they'll be much cheaper to build, require no launch costs, don't have to be perfectly airtight, and you get atmospheric pressure, water and oxygen for free. Also, in the case of war, plague and political tensions, ground-based semi-closed environments will be much less fragile and more survivable than sealed orbital tin cans.

Any space activists keen on setting up some of those first?

Comment: Re:It's not a miracle (Score 1) 204

by lennier (#47496417) Attached to: Apollo 11 Moon Landing Turns 45

Let's reach for the stars again!

Sure, but how?

Simple: wait 296,000 years for Voyager 2 to reach Sirius.

Oh, did you mean developing faster-than-light technology that will let us send probes to a star within a human lifetime, and with an energy output less than a supernova? We'll get right on that. First up, falsifying General Relativity, Quantum Mechanics and String Theory; shouldn't take too long...

Kidding aside, it seems like the 1960s Golden Age of GR was the last fun time; there really was a sense that we could engineer spacetime fabric Real Soon Now. Who knew what those wacky mesons did? Warp drive was just around the corner! But from the 1980s on, fundamental physics became the Science of Nope, You Can't Ever Have That. Science fiction in particular hasn't ever really recovered; it lives it its little parallel universe where the big future dreams of the 1930s live on.

Comment: Re:Outrage due to Censorship, not the test (Score 1) 219

by lennier (#47354819) Attached to: Facebook's Emotion Experiment: Too Far, Or Social Network Norm?

I talked to several (non-tech) friends about this, and they were more upset about Facebook "censoring" out posts than the emotional manipulation.

YES. This is exactly the problem.

Those of us who understand the tech already understand that Facebook's Newsfeed is not a 'dumb pipe' and that it runs an extremely opaque and proprietary filtering algorithm. We realise that a lot of posts get silently dropped; we constantly switch from 'Top Stories' to 'Most Recent' to try to counteract this. Many of us use Twitter instead and crosspost to Facebook because we know that Twitter tends to deliver all posts rather than silently screen them.

But non-tech-savvy people - our parents, grandparents, cousins, aunts - don't realise this. They think that the posts they see on the Newsfeed are exactly and only what their friends are saying. They think that if they see something with a thousand Likes, it's because their friends like it too. And why wouldn't they? To them, Facebook is a messaging service, not a media service. They know that TV and newspapers filter and select content. But they don't see Facebook as a newspaper. Their prior examples of messaging services are the telephone, post office, and then email - all three systems are ones that place a very high priority, almost a moral imperative, on the message always goes through unchanged without alteration or censorship. If they thought about it - which they generally don't - they'd expect that there was actually something in the user contract that specified this, because hey, isn't that the way things have always been? Isn't there something in the Constitution about freedom of speech? They don't realise just to what extent the Facebook terms and conditions say 'we reserve the right to hide posts from you, and not pass your posts on'. They don't realise that Newsfeed is far more like Rupert Murdoch paper than the Post Office.

That's the scandal here. It's shocking and it should be shocking to all your non-techy, non-cynical friends to see Facebook proudly talking about how they deliberately manipulate people's Newsfeeds to not actually be a representative sample of their actual friends' actual posts.

Keep the focus on that. The scandal is about censorship, free speech, and trust, not the esoterica of experimental protocols. It's important.

Comment: Re:It's really annoying (Score 1) 303

by lennier (#46701819) Attached to: OpenSSL Bug Allows Attackers To Read Memory In 64k Chunks

What languages is L4 written in?

The more relevant questions are "what is the size of the codebase of L4 written in an unmanaged language" and "is that unmanaged codebase small enough to mathematically prove its correctness" .

There is a reason why we layer systems on top of each other, and not just because we like cake.

Comment: Re:Yet again C bites us in the ass (Score 1) 303

by lennier (#46701771) Attached to: OpenSSL Bug Allows Attackers To Read Memory In 64k Chunks

What does managed code do that good C doesn't???

Managed code does one very important thing: it guarantees that elusive quality you've just named 'goodness'. (With respect to memory access, at least).

Goodness or otherwise of arbitrary unmanaged C code is a Turing-complete quality that, we've painfully discovered, cannot be reliably detected by either a compiler, a testing regime, or the entire planet's worth of expert C programmers given unlimited access to the code and up to two years time. That's how many coder-years? A lot.

Goodness of managed code? It has that quality. Period. And we can go on with our lives solving instead of creating problems.

Comment: Re:Gee, that's worse than no encryption isn't it? (Score 1) 303

by lennier (#46701223) Attached to: OpenSSL Bug Allows Attackers To Read Memory In 64k Chunks

If only they had written OpenSSL in Java instead of C!

Arguably all the recent security holes in Java are exactly because they wrote extensions and libraries in C/C++ and not in Java.

A real language - like, say, UCSD Pascal in 1978 can compile itself to its own virtual machine just fine...

But admittedly the resource requirements to host a system like that that are pretty steep - you'd need at least 128K of RAM. Still, I like to dream that one day....

Comment: Re:Um no (Score 2) 224

by lennier (#46597149) Attached to: Introducing a Calendar System For the Information Age

I don't think they're very concerned with easily-divisible numbers—4*7-day months and 13-month years!

13 months is a little annoying, yes; you have to split the months on week boundaries to make quarters. But we actually do have 13 lunar cycles in a year, so this naturally aligns the months with the real moon. And we keep 7 day weeks, which is a win both because we're used to our week, and because 7 days is a natural quarter-moon. And no more "30 days hath December..."

Thing is, a workable Earth calendar never is going to be evenly divisible by powers of 10, because it has to stay aligned with astronomical cycles which are subtly varying; even the Sun and Moon don't strictly align. So everything's going to be a bit of a juggle. Frankly, I think this is the best alternate calendar design I've seen in a long while.

Comment: Re:No, the problem is DVD should not be cheaper (Score 1) 490

by lennier (#46590163) Attached to: Are DVDs Inconvenient On Purpose?

The question you should be asking is why is streaming video so expensive that DVD (shipping little plastic discs around) is cheaper than sending bits over a wire?

Because it's the second stupidest deliberate misuse of computational capacity to artificially create digital scarcity since Bitcoins.

The correct way to distribute large files like movies online is to copy the bits as locally to the endpoints as possible, and cache them pervasively at all levels of the network. Nothing would need to be sent more than once down any given cable. It would be fast, cheap, make use of the Internet as it was designed to function, and give us near-unlimited bandwidth.

But that would mean that those bits don't become artificially scarce and can't be tracked and audited by the media companies for copy-protection purposes. So instead of copying, we stream them over and over and over again, generating terabytes of needless, duplicated data traffic, and creating huge bandwidth storms that suck all the capacity out of the Internet.

tldr: Video streaming is expensive because it was designed to be. It wasn't designed by or for you, and it doesn't benefit you.

"All my life I wanted to be someone; I guess I should have been more specific." -- Jane Wagner

Working...