Posts by Eddie Edwards

Re: I mentioned before

Not to mention the deliberate exploits the NSA put into quantum mechanics itself. You think you know what a transistor *really* does? Think again. They call it "spooky" action at a distance for a reason, you know.

Re: It is even slower than you say

>> bzip2 is 20% better than zopfli on enwik8 test

Wat? So Google have made a 100x slower compression that is already beaten by BWT from 20 years ago?

I guess they could add BWT and make it better, then? :)

Re: The law is an attempt to return to a bygone age

Well, exactly. That's what happens post-internet. Since we *are* post-internet, that's the situation. The genie is not going back into the bottle. Perhaps we should be more concerned with how we move forwards *given that* it's readily available, instead of trying to selectively maintain certain properties of a particular moment in time before the internet but after the invention of microfiche. (In fact, following your argument, I'm thinking microfiche may have been a really bad idea too. In fact, the very concept of an index is suspect.)

Re: Well reported.....

And of course the purpose of government is to take care of all the things no one cares about, to prevent it all going tits up. This is why you're forced to pay tax, rather than, say, just paying what you feel like towards things you care about today/this week/this month.

Re: Black Helicopters Herald in Programs and Pogroms, an Intelligence Coup?

First rule of El Reg: don't diss El Reg. I do so regularly and all my comments are moderated. They always seem to post my comments in the end, though, so I'm not sure why they bother. Probably to stop me getting into fights.

Second rule of El Reg: if you are going to criticize them, don't criticize them for being a red-top. They're a red-top on purpose. It's ironic - or at least it was. Now it's more self-parody. But, still, it's their identity, and you could no more have El Reg without that than you could have it without Orlowski's broadsheet-quality trolling.

Third rule of El Reg: amanfrommars has been here longer than you have. Even if you are amanfrommars - it's not clear it's even been the same person all this time. And I'm pretty sure his apparent schizophrenia is a put-on.

Re: How did they measure the 133Xe?

Yeah but the coincidence detector goes haywire every time the Heart of Gold flies past.

Re: No!

> > They'll say "Thanks very much" and promptly wrap it in proprietary licences so that they can gain the benefits and not the people who did the work."

The idea that your project fails *because people use it* is perhaps one of the strangest ideas to have come out of Stallman's head.

The idea that you have lost anything because other people are using digital copies of it in ways of which you do not approve is strange, to say the least (especially when you don't charge for it).

But the underlying idea that other people *should* only use your products in ways of which you approve - that this is something that is even desirable to enforce - is actually rather a disturbing notion. It's forcing your mind set onto others whenever you can acquire leverage to do so. It's certainly not remotely altruistic.

> > experience has shown that if proprietary companies get the opportunity they will abuse it and behave like parasites.

And yet, LLVM and Webkit have both seen major investment (including code submissions) from large companies like Apple, so it's hardly cut and dried that LGPL/BSD projects are this one way traffic from the kindly programmers in their bedroom to the big nasty corporations. The fundamental tenets of the "why" of the GPL are on very shaky ground if this kind of thing can occur spontaneously, wouldn't you say?

It invites the thought: maybe there is no "us" and "them" after all. Maybe there's just, like, people, man.

Re: Should be doable

It's fairly obvious that this doesn't work. The block size has to be larger than the amount of data needed to store the pointer-to-duplicate, otherwise deduplication is more expensive in disc space. So it needs to be at least say 128 bits. The likelihood of two random 128-bit blocks matching is astronomically small. Even if you have 2^64 blocks to choose from (250 million TB of data) the chance of a match is still only 1 in 2^64. You'd have to match a 250 mega-TB corpus with another 250 mega-TB corpus to get an expected saving of 16 bytes total.

Re: Kessel Run in less than 12 parsecs....

"If it's a fast ship?"

"You've never heard of the Millenium Falcon? I'm the guy who found a short cut to Kessel so it doesn't have to be fast."

Try again fanbois.

Re: Photoshop in Spaaaaace!

A quick Google shows the asteroid has a rotation period of 5-7 days, so it would not have rotated significantly during the flyby.

This isn't like those sedate fly-bys of Jupiter and Saturn. Wikipedia says the approach was taken at 10km/s relative velocity with a closest distance of 3km. The asteroid is 4km long. If you scale that to human units it's like driving past a house by the side of the road at 60mph, and trying to get a clear picture of the front. (If you scale to Voyager units it's like flying past Jupiter at 1/2 the speed of light!)

They did have quite a good camera on board though, so possibly it was the problem of rotating it fast enough to track an object at that speed. Maybe the camera mount doesn't rotate at all, only the craft. Lots of issues to solve to get that picture.

@ Chris Wareham

"I found myself thinking that I'd rather shovel pig shit for a living than have to work on a code base where ideas from that book had been used."

Unreal Engine 3 used ALL of them.

Then wrapped them in macros for the win.

Your pig shit metaphor is too kind.

Re: The writing was on the wall

"I've never heard of a 33mHz 386 chip — that'd be slow indeed!"

The great thing is it makes your HDD look like L1 cache.

Re: which way is the sun?

I think the idea is that the pin is a pin in the map, not a huge Space Needle type affair in the real world that you megalomaniacally control with your finger. Hence, the shadows would depend on your local light source, not on the light source at the mapped location. You'll notice they also fail to render the map in black when it's nighttime at the mapped location.

The real shocker here is that they're not using the front-facing camera to determine the location of the light sources where I am to make the drop shadow accurate even for that case. Clearly, it shows they have no attention to detail. I for one wish Apple would finally show some interest in Skeuomorphic interfaces.

@JDX

I'd say selection bias. The upvotes for the pro-RMS comments are due to a predeliction of people who follow the religion to be zealous about it. GPL people care enough to upvote that comment; non-GPL or anti-GPL people don't.

The existence of selection bias makes it hard to accurately gauge the demographic from the comments section. But the upvotes obviously prove the presence of a pro-GPL/RMS contingent which was either absent or less active a few years ago.

As for "pro-FOSS", that could mean pro-BSD and anti-RMS. FOSS is a rather wooly term and I'm not sure anyone is *against* having open-source software available :)

Re: Random levels required

Lee, your two posts on this are remarkable in that they give a very good sense of why this is a very interesting problem and why it's an interesting game, yet you are simultaneously telling us how pastiched and boring both the game and the problem of playing the game are!

Re: Fake nonsense statistics based on flawed samples

50% of revenue going to 25 developers is not "flawed samples". You can't accidentally get a result like that through a few rounding errors. Rather, it's clearly indicative of a power-law distribution, and as such is completely expected in a system where there is no hard maximum value for revenue.

As for the idea that if there wasn't a chance of success, people wouldn't try ... see "The National Lottery". (There's an extreme power law - consider the proportion of payouts which go to ~52 people each year.)

But there is a very silver lining here. The key fact is, the *other* 50% is going to the other developers. Big software houses are not getting all the profits. The big software houses are getting half the profits and leaving half on the table for smaller fry. Half the profits not going to the big boys is very good news indeed. Compare this with the movie industry, the traditional computer game industry, or the recorded music industry.

It's still not brilliant news for everyone though. If it is a power law distribution, then the next 25% goes to the next 25 developers, the next 12.5% to the next 25 developers, and so on - so 94% goes to the top 100 developers. Still, even the remaining 6% of the App Store pie is a ton of money. You can in theory figure out exactly where you need to be in the pecking order to get $X revenue per month. And it's high, but not that high, which is why we do see many individuals being financially self-supporting via App Store. But the number of people who have failed to even make $50 back must be 10 or 20 times higher than this.

Ha, I've seen it over many years but it's only your post which made me realize why it's used.

Re: 7.8336MHz?

Thanks, I thought someone would know :) So, not an integer frame rate then (I make that 60.147Hz). So the precise clock rate still makes little sense, unless for some reason 15.6672MHz oscillators were commonplace and 15.6288MHz (for 60.000Hz) oscillators were hard to find!