SG1 wanders around the place a bit, but it regularly does very 'hard' sci-fi. Another common theme of theirs is to follow through on the human repercussions of the hyper-tech, including the inevitable stupid mistakes. For anyone who's not seen the series, or had the misfortune to see average episodes, (the first season in particlar turned a lot of people off) try these. Each episode is about 20-30Mb. ASF files are DivX.
1.18 - Tin Man Fantastic episode. Can't say any more, or I'll spoil it.:)
2.16 - A matter of time. It's not often that General Relativity is a major plot point. (Ignore the glaring stellar physics mistake made in the first minute.) The ending is slightly lame, though.
2.21 - 1969 There's always the obligatory accidentally-sent-back-in-time episode. What matters is how you handle it.
3.4 - Legacy Great mystery. Gets really spooky in parts. Has one of the classic SG1 'whoops' scenes. Repeat after me: biological weapons are BAD.
3.6 - Point of View They did the typical parallel universe episode in season 1. But of course, the door opens both ways...
"I believe Alex Chiu's insights on life, the
universe, and everything are every bit as
scientifically valid as those of the
(late, lamented) Douglas Adams" - Roblimo
That REALLY hurt. In fact, I don't think I've felt this angry over such a simple statement in a while. I'm not even sure exactly why. I'll try to work that out in a moment, but in the meantime, I'll just say I think that was in extremely poor taste.
On the surface, there are some similarities. Douglas did make fantastic extrapolations from scientific principles, mixing them with strange ideas to produce Infinite Improbability Drives, Somebody Else's Problem fields, etc.
On the other, Douglas was well aware he was writing fiction. In his non-fiction work he showed quite clearly his firm (though non-expert) grasp of evolution, biology, and even computer science. He took an interest. He tried to learn, to understand, everything around him on it's own terms.
Alex Chiu has a fixed agenda. He has a particular view of the world which he strongly endorses at every opportunity. His research is designed entirely to support his thesis. He seems to have no interest in anything outside this.
The fundamental difference is that Douglas was NEVER SURE of anything. He would sieze new ideas and play with them for weeks (much to the annoyance of his publishers) just trying to cope with them. He was perpetually astonished by the world, and science, and his writing tended to be an expression of this. He communicated that wonder and astonishment to everyone.
Douglas asked us to question everything with wonder. Alex asks you to believe his truth with certainty. Which is closer to the scientific method?
So, No. I think Douglas' insights were much more scientifically valid than Alex's.
This only just occured to me, and I've been awake for an awful long time now, but I would seem that if you're invoking fair use for journalistic reasons in a situation where the media counts as 'evidence' of some kind, then the quality of the transfer is paramount.
To illustrate, let's pretend that the Rodney King beating was filmed with the video equivalent of MiniDisc. If the device does not allow a full-quality digital backup to be made of the recording, then vital information (the identity of the officers, say) would be lost in the blur of a degraded signal.
Or worse, the lack of quality could be interpreted as tampering, or be used to conceal it.
Documents important enought to be guaranteed distribution under fair use must, almost by defninition, be deserving of full quality rendition.
I'll restate. The chance of matching a 100 bit string is so remote - you need to search through so much of PI - that to simply record WHERE in PI you found the match would require about 99 bits just to store a number that big.
Incidentally, 2^99 is about 10^97, which is up in the number-of-subatomic-particles-in-the-universe range. If you actually try to compute 2^99 bits of PI, you'll be here long past when the sun burns out. You have to do that both during the compression and the decompression steps BTW. But I digress...
Any number you store requires at least a little metadata about it - it's length, usually - which will require more than 1 bit. Thus, there is no saving after all.
Bottom line. You cannot compress truly random data. That's the definition. If you can compress it, it wasn't random.
Since a couple of people have asked the same thing, I'll have a go. My mathematics is reasonable, but I'm not a pro. IANAM.
PI isn't actually random, and will skew this analysis in ways that doubtlessly earned someone a PhD, but let's ignore that for the moment. Let's pretend PI is an infinite-length random bitstream. It doesn't matter much.
Being infinite in length, it will therefore (eventually) contain any finite length random bitstring that you care to look for.
The chance of finding a match at any location is one in 2 to the power of the length of the bitstring. So, the chance of finding a 100 bit match at any index is 2^100.
You will therefore find a match, ON AVERAGE, at a distance of about 2^99 bits into PI.
Hurrah! you say. We only need a 99 bit index to represent a 100 bit number. We've made a saving!
First, let's just diverge slightly and apply a sanity check for why this can't possibly be right. If that logic holds, then we can compress any N bit number to N-1 by looking for it in PI. Therefore, we can compress the 99 bit index to a 98 bit index, then to 97... down to 1 bit. Something's wrong.
The issue is that you get a 99 bit index ON AVERAGE. Sometimes it will be 101 bits. Sometimes it will be 98. There's even a small chance (exactly 2^100) it will be right at the beginning. So, you can't just have a fixed 99 bit block to store the index, you need another number to store the length of the index number.
Since the number "99" requires 7 bits of storage, your average case will be to represent a 100 bit string as a 106 bit index+length code.
So, actually, you end up expanding the data, if the law of averages holds.
There's no way around this, no matter how clever you get.
If you say "we'll use a 99 bit block only if it fits, otherwise we'll just store the original 100 bit number" then you need at least a 1 bit switch to indicate this.. so on average, half will be 99+1 bit blocks, and the other half will be 100+1 bits. 100.5 bits per original 100 bit block.
No, Huffman coding the index length doesn't help either. But if you asked that question, you probably already knew the answer.
Of course, you might get lucky on individual cases. That's always a possibility when dealing with truly random data. But then, that's not compression, that's gambling.
If you really want to learn about this stuff, read Claude E. Shannon's work on Entropy and coding theory. (also Turing, Huffman, and Godel) Doing so will also give you the grounding necessary for thermodynamics and quantum computing.
A travelling programmer came to the cubicle.
"I seek Elegance" he said.
"Why have you come here?" asked the Master.
"Because I have read much of your code, and it is elegant. I have come to
find the source of it, so that I may write good code also."
"Then you have come to the wrong place." said the master.
"But are you not the writer of the code?" said the programmer.
"How did you know the code was elegant?" asked the master.
The programmer was enlightened.
I'd come to pretty much the same conclusions over the last few months. I've been talking to engineer friends about the machines that sinter metal powders with lasers, the UV-sensitive polymers, and the starch bubble-jet printer, and the possible combination of them with the various plastic electronics components still being developed.
I applied Moores Law (probably not directly relevant, but a good bechmark nonetheless) to the current state-of-the-art, which is a washing machine sized device that can sinter plastic powder. A place here in town will take your CAD files and return a plastic part for about US$180 a pop.
I calculated we'll have microwave-sized 'fabbers' as a common household appliance in a little less than 15 years.
That means us early apopters will have expensive 'hobby' kits in just less than ten.
The main action will remain in commercial prototyping up until then. It will probably be a common commercial practise in about five years.
Right now the only users are the commercial and educational 'early-adopters' with specific needs that are filled by the fairly primitive tech we have now. Automated milling machines are currently a far superior tech if you want to do metal, for example.
Of course, the very existence of these machines is likely to push the curve forward by an unpredictable amount.
Just consider: fifteen years from now, well look back on all those jokes about ordering things over the internet, ("how do you fit it down the wires! Ha, ha!") and not get it.
The main obstacles are as follows:
* Sintering (melting powders together) has precision limits caused by heat transfer in the material.
* Light-sensitive polymers are still rather nasty chemicals. They're also quite brittle.
* Plastic electronics need to become available in dye/powder form. This will take a while. (And they will never beat silicon for high-performance tasks)
* You need infrastructure to ship the raw materials. They don't just magically materialize in the print-head, you know.
* Recycling all these things is going to become an issue.
But, yes. 'Fabbers' are another great step down the path. Open Hardware, here we come. (And the advantage is, it's easier to design a plastic toy than your average piece of software)
If you want a good historical analogy, you can't go past printing. From Gutenberg, to desktop publishing, to the web.
I've done a little in this area. Here's some suggestions:
* Resist the urge to stream _everything_. Sometimes it's just better to put the next lesson into a.zip file and mail it to them the day before. You don't want lessons being disrupted because someone's modem drops or the webserver crashes or it's just a slow net day. It also leaves bandwidth for the important stuff.
* Videoconferencing isn't as useful as it sounds. Apart from the bandwidth requirements, you're lucky if what you want to look at is framed properly without multiple cameras. Everyone will want it, however (It's one of those buzzwords, you see) so say it's planned for 'phase 2'.
* Audioconferencing, however, can be extremely useful. If you're going to do it, never forget the possibilities inherent in the humble telephone. Especially as a backup when the modem explodes.
* Work hard on making the audioconferencing work well. Ideally, you want a 'party line' situation, where students can hear the teacher, and can also talk to other students and hear the questions they ask. Also, pay attention to the hardware, which is usually the weak link. Get headsets with little boom mikes and software which 'normalizes' the volume.
* If computers are a central part of the lesson, investigate remote-screen viewing software like VNC. It lets you see that students are coping, help them out when they get stuck, and the knowlede that the teacher is also looking over their shoulder tends to keep students focussed. It doesn't use a lot of bandwidth, either.
Think of welfare not as something which keeps "bludgers in booze money", but something that stops them from having to break into your house and kill you in order to survive.
Cut welfare, and that's what you get. Take away the safety net, and see what the cost to society is when tens of thousands of people start hitting rock bottom. Violent crime increases. You either pay for people to not do anything, or you pay for the consequences.
Unemployment is inevitable. We're been running at about 10% for as long as I can remember, most of that in youth unemployment. Given that we need less and less people every year to do the same work (all that 'efficiency' we keep saying is a good thing) the situation is only going to get worse. Oh, look what we're doing to Centrelink! That'll help!
All the stupid government make-work programs in the world aren't going to help this, either. We should make education free again, so they can go back into the universities where they belong:-)
Seriously though, spending 10 years trying to get an Arts degree (ie: surfing and drinking beer on Austudy) is better than breaking and entering because there's no jobs, no money for this weeks' rent - or food - and no future.
As for your cretinous attitude to the Arts... it's on par with your misunderstanding of politics.
Sadly, the Australian government DOES listen to Microsoft. Every time Bill Gates comes out here, politicians fall over themselves to get photographed with him, our sad little prime-minister included.
It might also have something to do with MS's heavy involvement with the media over here. NineMSN particularly. (Think MSNBC) Politicians are always nice to people with significant media influence.
And partly it's the tired old "He's the richest man in the world, so he must be right."
Not that Microsoft doesn't ALREADY have a significant corporate presence here. They may not do much R&D, but their Sales and Certification departments are pretty big.
[I posted this in response to another comment, but I think it's relevant here too.:-)]
While I agree that the government here - on IT issues especially - couldn't find it's own arse if it used both hands and a map, I think Microsoft's suggestions are uniformly self-serving and would actually make the situation worse.
Our vaunted 'deregulation' has turned Telecom from a slothful but competent beauracracy into an agency that gouges more money from Australians than can be imagined, (biggest profits in australian corporate history) while sacking most of the technicians who actually run the system. Remember; they only started charging that $0.17/Mb in 1995 once they'd wrested control of AARNet from the vice-chancellors committee. As for the situation in the bush...
Frankly, what this nation needs is not more corporate investment and control, but:
* Nationalization of the telecom infrastructure (yes, you heard me) and deregulation only of services and support. Same for electricity, water, and all basic infrastructure; which is frankly why governments are supposed to exist.
* Significant government investment in education and basic research. Again.
* Sane and sensible policy on issues like digital signatures, privacy, and datacasting. (Die! Alston, Die!)
* Use of government clout to introduce new technology once past the incubator stage, which is where we have traditionally fallen down.
Of course, none of that is going to happen. I despair.
While I agree that the government here - on IT issues especially - couldn't find it's own arse if it used both hands and a map, I think Microsoft's suggestions are uniformly self-serving and would actually make the situation worse.
Our vaunted 'deregulation' has turned Telecom from a slothful but competent beauracracy into an agency that gouges more money from Australians than can be imagined, (biggest profits in australian corporate history) while sacking most of the technicians who actually run the system. Remember; they only started charging that $0.17/Mb in 1995 once they'd wrested control of AARNet from the vice-chancellors committee. As for the situation in the bush...
Frankly, what this nation needs is not more corporate investment and control, but:
* Nationalization of the telecom infrastructure (yes, you heard me) and deregulation only of services and support. Same for electricity, water, and all basic infrastructure; which is frankly why governments are supposed to exist.
* Significant government investment in education and basic research. Again.
* Sane and sensible policy on issues like digital signatures, privacy, and datacasting. (Die! Alston, Die!)
* Use of government clout to introduce new technology once past the incubator stage, which is where we have traditionally fallen down.
Of course, none of that is going to happen. I despair.
Sadly, I have to agree that Academic R&D is going downhill. CSIRO badly needs more funding, and the situation with Universities is going critical. (Literally, with the 'academic bribery' allegations currently flying around like neutrons in a reactor)
I agree that a lot of good R&D gets done interally by companies, possibly even more in total (I myself would probably be considered such) but you state exactly the problem: It gets exported. Or quietly dropped.
At least with academic research, the tradition of publishing everything means that whatever reasearch is actually done adds to the total pool of knowledge, if only to show "this doesn't work". Too much corporate research never sees the light of day, especially the failures, which are often the most important.
I know this is a generalization, but too much corporate 'R&D' crosses the line to 'Applied', in my opinion. Exactly what the academics are being pushed towards.
I really should read the whole 'Accellerating Innovation' paper before commenting on details, but I can give a first-glance perspective:
On Bandwidth:
Australia was significantly behind the US in terms of broadband until about six months ago. Cable and DSL have arrived here with a vengance! $60 a month will get you 400kbit/s DSL. Really. The technical implementation has been rock-solid: none of the problems the US has been having. There are still political issues with the 'duopoly' telco situation, but the ACCC (Australian Competition and Consumer Comission) should do a good job of sorting out the more draconian stupidities in their 'Acceptable Usage Agreements' if they try to enforce them.
(The ACCC is way cool. They're currently having a spat with the MPAA over DVD regional coding.)
Home users here are in the best position they've ever been in. The biggest problem is that Telstra keeps running out of cable modems.
Academic and Business users are, however, in a position of paying rather more for interet access than is really sane. Telstra and Optus C&W both charge about $0.17 per megabyte for inbound traffic, (yes, you heard right) something which deregulation should have put an end to. But Optus likes the money too. Things should change with the recent landing of the Southern Cross Network pacific fibre.
Microsoft wants more broadband in Australia because they've targeted us as one of the prime test markets for.NET. Make of that what you will.
As far as accessibility goes, anyone who wants net access can get it. It's in every library, government department, and medium to large business. ISP's and net cafe's abound. The only people who don't have access are those too lazy to get off their fat arse and walk down the street.
On Taxes:
Yes, Australia has fairly high income and corporate taxes. We have correspondingly higher 'socialist' programs like decent state-sponsored Education, Medical care and Welfare. Australians are, on the whole, happy with this situation. Similar to Canada, in many respects. Multinationals don't like high taxes, though.
On R&D:
There is a cultural difference in what R&D means. We still have a very strong Academic research tradition: most really innovative stuff in Oz gets done by the CSIRO (commonwealth science and industry research organisation) and the universities. I think this is fine. Corporations suck at real R&D. I think Microsoft are just complaining they can't twiddle their thumbs inventing things like "Microsoft Bob" and get massive tax rebates for it.
Besides, what most US companies would call R&D, we call 'getting shit done'. We're an emminently practical people in many ways.
Failure to create local centres of excellence:
Rubbish. See above. The main issue seem to be the existing centres of excellence contain people who work outside the corporate domain and are, therefore, fairly unpoachable. ie: No talent pool waiting for Microsoft to snaffle them all.
On Piracy:
Australia has one of the highest rates of 'Piracy' in the western world. And literacy too. We are significant users of information products, even if we can't quite afford it.:-) Most of that 'piracy' happens on an individual level. Software swapping with friends, etc. Small and 'backyard' businesses run a lot of 'grey' software. Large companies tend to be utterly legit, though.
There are some historical reasons for that, mostly due to semi-monopoly locks on imports of information products (the mentioned "parallel import restrictions") like software and music that kept the price artificially high for a long time. These have been changed recently, and will change again, since anyone with a net connection and credit card essentially ignores it.
A skills shortage:
"Australia faces many specific hurdles to train and retain skilled people."
Duh!! You know why? Because of the famous 'brain drain'. US/UK companies tend to make our best and brightest offers they can't refuse. We'd have better people if Microsoft stopped poaching them all.
On Capital:
Australia doesn't have anywhere near the level of Venture Capitalism that exists in the states. (but who does?) Yes, this does hurt us. He has a point here. But, if microsoft want to fix it, all they need to is start investing themselves. A billion US dollars goes a long, long way here. Microsoft has no lack of access to capital. I think what they mean is; there's no easy pool of money for Microsoft to suck from.
Technology security:
I can't think of any Oz sites that have been DoS'd, or even defaced. (I'm sure there's been some. Nothing to the level of Yahoo or Ebay. Mostly a US phenomenon) All major banks in Australia now offer internet service, and I've never heard of any problems. What has the writer been smoking?
Summary:
Yes, as many people would expect, Microsoft are complaining about local factors that tend to favor individuals and small businesses over large multinationals like them. There are one or two decent points in the dross, specifically those about lack of telecom deregulation and capital availability.
The gist of the complaint seems to be "give us more money or we won't invade your nation".
h'Wha? Remember, I've done this. Transparency works. It's easier to show than explain, so I've put up some screenshots here:
http://www.ecn.net.au/~orinoco/ideas/grid3d.jpg
Complaints about visual bugs can go straight to/dev/null. This was written in two weeks with absolutely no prior experience of J3D. I say again: Java3D rocks. It will rock even more so when they finish it:-)
"How is navigating a hierarchy with "hidden" objects/tasks/whatever better than a linear list of tasks/objects/whatever if the screen has plenty of space to display the tasks in a linear manner"
It's not. I didn't say anything about hierarchies or linear anythings. Take a look at the shots. The system I built actually works more on a hyperlink principle.
For anyone interested, it failed for two reasons:
1. Nobody had the hardware to run it.
2. Users kept getting lost.
Whoops. I've been doing too much Java3D.:-) When I said, "20% transparent", I really meant "20% opaque". You want to be able to see through about five layers of objects before they get completely occluded.
And don't go overboard on the lighting. Diffuse is enough. Specular lights are just distracting, even if they look cool.
Having recently spend two weeks building an ultimately unsucessful 3D database visualization system, (Java3D rocks!) I can tell you it's a LOT harder than it sounds, a bit like speech recognition.
The main problem is that there is no consistent paradigm for 3D interaction... no equivalent of a desktop metaphor. Users find themselves lost in space. And such systems are hard to interact with properly with a 2D screen and mouse - the missing degree of freedom create a 'modal' system that cannot be intuitive.
But, if you have to do it, here are Orinoco's tips:
1: Make everything about 20% transparent. You can't work with half the environment hidden behind the other half.
2: View control is the key. Don't make the user have to spin and rotate. Let them pick objects of interest, and then move the camera to a good view of it.
3: Don't try to model a complete 3D environment. Instead, make it "2.5D", with the extra dimension used to express an intrisic scalar quantity rather than a spatial one. eg: A 2-D scatterplot, but each point is instead a bar who's height indicates something.
4: Create a 'groundplane'. Stick to stacking things above this.
5: A 'recursive boxes' scheme, with whole new scenes hidden inside pickable objects works well. (A folder metaphor, if you will)
Just consider the most effective 3D application yet - 3D modelling. Even with a perfect 1:1 correspondence between the visual representation and the underlying model, it still takes experts to manipulate the interface.
Frankly, I think 3D interfaces have to wait until we have cheap and available 3D input devices.
Finally, there's a lot of research that has been done (SIGGRAPH, to name a source) that you would be silly to ignore.
I'm sorry. I can't buy into Froomkin's rant that ICANN is Evil because it's not under the control of the US Constitution. Hello? It's the INTERNET. Whatever process they come up with will conflict with some country's Constitution or ideology. That's why they were made independant... so they could function seperately. Duh! Please don't harp on how they're organised. It makes you look like a redneck to us 'Non-US Internet Users'.
He's got some legitimate concerns about the current agenda of ICANN and corporate influence, but I'm all for new TLDs... make hundreds! Use the entire dictionary, in fact! (Better than picking some poxy subset..biz? Ew!) Then the squatters can't get ALL of them, surely? As far as the commercialism goes, I'm resigned to watching sleazy companies try to charge me fifty bucks to put a tuple in a database, and somehow get it wrong.
Lastly, DNS is fleeting. It is unlikely to survive for more than five to ten years, now. It's incredibly replaceable, not only with parallel DNS systems run by muliple 'authorities', (would make DNS a little more complicated to admin, but with some simple tools...) or you can simply turn your network nameserver into a gateway to Whatever Comes Next(tm), without having to touch any old clients, as long as the names still map to something.
There are numerous possibilities for totally distributed non-unique locator systems. Go talk to the mathematicians about Simulated Annealing, and the cryptologists about identity certificates. Hell, this latest paper on vertex saturation might be quite useful, too.
Finally, we should learn the lesson of centralism. DNS was a centralized point of power. It has been corrupted. So ends the lesson.
I will do this by means of character play: Corporate Server: Hi there! Welcome to GlobalCorp! Modified Mozilla Client: Yeah, hi. S: Please auto-download your details now. C: I don't want to. S: But you have to! C: I really don't want to. S: Then you can't have the page! Nyah. C: Oh, all right then. Bill CLinton. 1400 pensylvania avenue. Female. Guatamala. Sheep. S: Beep. Thank you. You may now read the page. C: [aside] Suckah. What P3P does is allow the machine to do all this for you! Transparently even! With no more of those stupid on-line forms. Mmmmm. Technology==Good. ~ Orinoco
To those in the know, this news seems almost expected. All scientists have curiosity, often triggered by what it right in front of them for much of the time. Therefore, it is only natural that some Australian scientist[1] should be sufficiently intrigued by strange bubbles in his beer, to mount a research effort of this magnitude. Helped, of course, by his supportive peers.
This is simply another step along the road to the GUB. (Grand Unified Beer) Slowly, we assembling a body of theory that will let us unify the four beers (Carlton, Swan, Guinness, and XXXX) at high energies. At Board length, some brewers theorise that 'Superbubbles' will form, and if properly chilled, it condenses into 'Superbeer'. Alas, it may be necessary to explode a nearby star to obtain the required energy. Ah well. No-one said good beer was easy. Please brace yourselves...
After this, we plan to turn our attention to Barbeque technology. We're impressed by the lighting times of the American Liquid Oxygen/Briquette BBQ, but have found it difficult to cook with, mostly since the steaks and sausages reduce to a charchoal impurity in the melted mass of the plate. We feel meat should be slightly rarer. Currently, we're tinkering with a solar furnace BBQ which is capable of cooking snags eight at a time at a distance of 30 metres.
In the ultimate model, orbiting solar mirrors are used to call down fire from the sky[2] into either permanent-emplacement BBQ recievers, or a portable 'campfire' version with GPS targeting system. Not to be used indoors.
Viva Australia! Mate.
[1]We are often defined by our environment. That which surrounds us, we are best at. Some prime examples where Australia still leads the world are; Solar Energy (due to a sun that will, given time, kill you with sheer brightness); and in developing antitoxins to nature's amazing array of very poisonous things. (Almost all of which live here, often in your boots)
[2]Clarke's law: Any technology sufficiently advanced is indistinguishable from magic)
Feynman is up there. Pauli too. Strangely, I'd also add Ghandi to the list, perpetrator of the ultimate social hack, someone who made the future happen by living it just as hard as he could.
There are so many. Descartes. Bose. Schrodinger. Planck. Euler. (or was he too early?) Mandelbrot. Tesla. Fermi. Escher. Godel.
More recently and closer to home, The Woz, John West of DEC, (see: Soul of a new Machine) Ritchie and Kernigan. David Braben. Tim Berners-Lee. Bill Joy. Jaron Lanier. And of course, Donald E Knuth.
John Postel. Lest we forget. The whole IETF and W3C, for that matter.
Srinivatha Ramanujan.
Damn, I hope I spelled that right.
Without a doubt the greatest mathematician ever. His story is one of brilliance and tragedy. From a single old mathematics textbook, he extrapolated the entire of western mathematics, and went a hundred times further. Superstring theory depends on some of his math. And there are still notebooks full of his work which haven't been fully understood. He died young. Too young. Such loss is inexpressible.
Another brilliant but tragic geek is Taniyama, of the Shimura-Taniyama conjecture (again, I hope I got that right) which was recently finally proved, and formed the keystone of Fermat's last theorem[1]. He committed suicide.
For my part, I'd take Newton off any list due to his very ungeeky behavior outside physics[2]. Shockley is another I waver on. He invented the transistor, but wasn't very nice about it.
There's too many. The problem is we all stand on each others shoulders, all the way back to Aristotle. I'm reminded of a french mathematician priest who's enduring contribution to math was his letter writing, corresponding with dozens of mathematicians like Fourier, passing ideas and problems from one to the next. I don't remember his name. As a rule, most people don't either. I suspect the greatest geek of the century is one who made their mark in the same quiet, unobvious way. Their ideas so natural and obvious after the fact that no-one even wondered where they came from.
I say, leave the number one spot empty. Let it stand unfilled because we'll never be able to give it to the person who deserves it most.
~ Orinoco
[1] It's basically a mapping system which joins together two huge areas of mathematics. The reason it's important is that impossible problems in one domain can now be 'ported' across to another, solved there, and ported back.
[2] He was a cruel and vindictive man, who went out of his way to destroy the reputations of rivals. Brilliant, yes, but a nasty piece of work. I'm not surprised that people sat in trees and dropped fruit on him.:-)
Slashdot Mirror
SG1 wanders around the place a bit, but it regularly does very 'hard' sci-fi. Another common theme of theirs is to follow through on the human repercussions of the hyper-tech, including the inevitable stupid mistakes. For anyone who's not seen the series, or had the misfortune to see average episodes, (the first season in particlar turned a lot of people off) try these. Each episode is about 20-30Mb. ASF files are DivX.
If I had to choose just one, "Window of Opportunity" would be it.
"I believe Alex Chiu's insights on life, the
universe, and everything are every bit as
scientifically valid as those of the
(late, lamented) Douglas Adams" - Roblimo
That REALLY hurt. In fact, I don't think I've felt this angry over such a simple statement in a while. I'm not even sure exactly why. I'll try to work that out in a moment, but in the meantime, I'll just say I think that was in extremely poor taste.
On the surface, there are some similarities. Douglas did make fantastic extrapolations from scientific principles, mixing them with strange ideas to produce Infinite Improbability Drives, Somebody Else's Problem fields, etc.
On the other, Douglas was well aware he was writing fiction. In his non-fiction work he showed quite clearly his firm (though non-expert) grasp of evolution, biology, and even computer science. He took an interest. He tried to learn, to understand, everything around him on it's own terms.
Alex Chiu has a fixed agenda. He has a particular view of the world which he strongly endorses at every opportunity. His research is designed entirely to support his thesis. He seems to have no interest in anything outside this.
The fundamental difference is that Douglas was NEVER SURE of anything. He would sieze new ideas and play with them for weeks (much to the annoyance of his publishers) just trying to cope with them. He was perpetually astonished by the world, and science, and his writing tended to be an expression of this. He communicated that wonder and astonishment to everyone.
Douglas asked us to question everything with wonder. Alex asks you to believe his truth with certainty. Which is closer to the scientific method?
So, No. I think Douglas' insights were much more scientifically valid than Alex's.
Yes, I'm biased.
And I miss him terribly.
This only just occured to me, and I've been awake for an awful long time now, but I would seem that if you're invoking fair use for journalistic reasons in a situation where the media counts as 'evidence' of some kind, then the quality of the transfer is paramount.
To illustrate, let's pretend that the Rodney King beating was filmed with the video equivalent of MiniDisc. If the device does not allow a full-quality digital backup to be made of the recording, then vital information (the identity of the officers, say) would be lost in the blur of a degraded signal.
Or worse, the lack of quality could be interpreted as tampering, or be used to conceal it.
Documents important enought to be guaranteed distribution under fair use must, almost by defninition, be deserving of full quality rendition.
No.
I'll restate. The chance of matching a 100 bit string is so remote - you need to search through so much of PI - that to simply record WHERE in PI you found the match would require about 99 bits just to store a number that big.
Incidentally, 2^99 is about 10^97, which is up in the number-of-subatomic-particles-in-the-universe range. If you actually try to compute 2^99 bits of PI, you'll be here long past when the sun burns out. You have to do that both during the compression and the decompression steps BTW. But I digress...
Any number you store requires at least a little metadata about it - it's length, usually - which will require more than 1 bit. Thus, there is no saving after all.
Bottom line. You cannot compress truly random data. That's the definition. If you can compress it, it wasn't random.
No.
> And hence *on average* you win the competition.
Only if you can devise a 1 bit decompression algorithm. Good luck.
Since a couple of people have asked the same thing, I'll have a go. My mathematics is reasonable, but I'm not a pro. IANAM.
PI isn't actually random, and will skew this analysis in ways that doubtlessly earned someone a PhD, but let's ignore that for the moment. Let's pretend PI is an infinite-length random bitstream. It doesn't matter much.
Being infinite in length, it will therefore (eventually) contain any finite length random bitstring that you care to look for.
The chance of finding a match at any location is one in 2 to the power of the length of the bitstring. So, the chance of finding a 100 bit match at any index is 2^100.
You will therefore find a match, ON AVERAGE, at a distance of about 2^99 bits into PI.
Hurrah! you say. We only need a 99 bit index to represent a 100 bit number. We've made a saving!
First, let's just diverge slightly and apply a sanity check for why this can't possibly be right. If that logic holds, then we can compress any N bit number to N-1 by looking for it in PI. Therefore, we can compress the 99 bit index to a 98 bit index, then to 97... down to 1 bit. Something's wrong.
The issue is that you get a 99 bit index ON AVERAGE. Sometimes it will be 101 bits. Sometimes it will be 98. There's even a small chance (exactly 2^100) it will be right at the beginning. So, you can't just have a fixed 99 bit block to store the index, you need another number to store the length of the index number.
Since the number "99" requires 7 bits of storage, your average case will be to represent a 100 bit string as a 106 bit index+length code.
So, actually, you end up expanding the data, if the law of averages holds.
There's no way around this, no matter how clever you get.
If you say "we'll use a 99 bit block only if it fits, otherwise we'll just store the original 100 bit number" then you need at least a 1 bit switch to indicate this.. so on average, half will be 99+1 bit blocks, and the other half will be 100+1 bits. 100.5 bits per original 100 bit block.
No, Huffman coding the index length doesn't help either. But if you asked that question, you probably already knew the answer.
Of course, you might get lucky on individual cases. That's always a possibility when dealing with truly random data. But then, that's not compression, that's gambling.
If you really want to learn about this stuff, read Claude E. Shannon's work on Entropy and coding theory. (also Turing, Huffman, and Godel) Doing so will also give you the grounding necessary for thermodynamics and quantum computing.
A travelling programmer came to the cubicle.
"I seek Elegance" he said.
"Why have you come here?" asked the Master.
"Because I have read much of your code, and it is elegant. I have come to
find the source of it, so that I may write good code also."
"Then you have come to the wrong place." said the master.
"But are you not the writer of the code?" said the programmer.
"How did you know the code was elegant?" asked the master.
The programmer was enlightened.
Ha. "Fabbers". Good name.
I'd come to pretty much the same conclusions over the last few months. I've been talking to engineer friends about the machines that sinter metal powders with lasers, the UV-sensitive polymers, and the starch bubble-jet printer, and the possible combination of them with the various plastic electronics components still being developed.
I applied Moores Law (probably not directly relevant, but a good bechmark nonetheless) to the current state-of-the-art, which is a washing machine sized device that can sinter plastic powder. A place here in town will take your CAD files and return a plastic part for about US$180 a pop.
I calculated we'll have microwave-sized 'fabbers' as a common household appliance in a little less than 15 years.
That means us early apopters will have expensive 'hobby' kits in just less than ten.
The main action will remain in commercial prototyping up until then. It will probably be a common commercial practise in about five years.
Right now the only users are the commercial and educational 'early-adopters' with specific needs that are filled by the fairly primitive tech we have now. Automated milling machines are currently a far superior tech if you want to do metal, for example.
Of course, the very existence of these machines is likely to push the curve forward by an unpredictable amount.
Just consider: fifteen years from now, well look back on all those jokes about ordering things over the internet, ("how do you fit it down the wires! Ha, ha!") and not get it.
The main obstacles are as follows:
* Sintering (melting powders together) has precision limits caused by heat transfer in the material.
* Light-sensitive polymers are still rather nasty chemicals. They're also quite brittle.
* Plastic electronics need to become available in dye/powder form. This will take a while. (And they will never beat silicon for high-performance tasks)
* You need infrastructure to ship the raw materials. They don't just magically materialize in the print-head, you know.
* Recycling all these things is going to become an issue.
But, yes. 'Fabbers' are another great step down the path. Open Hardware, here we come. (And the advantage is, it's easier to design a plastic toy than your average piece of software)
If you want a good historical analogy, you can't go past printing. From Gutenberg, to desktop publishing, to the web.
Can't wait.
I've done a little in this area. Here's some suggestions:
.zip file and mail it to them the day before. You don't want lessons being disrupted because someone's modem drops or the webserver crashes or it's just a slow net day. It also leaves bandwidth for the important stuff.
* Resist the urge to stream _everything_. Sometimes it's just better to put the next lesson into a
* Videoconferencing isn't as useful as it sounds. Apart from the bandwidth requirements, you're lucky if what you want to look at is framed properly without multiple cameras. Everyone will want it, however (It's one of those buzzwords, you see) so say it's planned for 'phase 2'.
* Audioconferencing, however, can be extremely useful. If you're going to do it, never forget the possibilities inherent in the humble telephone. Especially as a backup when the modem explodes.
* Work hard on making the audioconferencing work well. Ideally, you want a 'party line' situation, where students can hear the teacher, and can also talk to other students and hear the questions they ask. Also, pay attention to the hardware, which is usually the weak link. Get headsets with little boom mikes and software which 'normalizes' the volume.
* If computers are a central part of the lesson, investigate remote-screen viewing software like VNC. It lets you see that students are coping, help them out when they get stuck, and the knowlede that the teacher is also looking over their shoulder tends to keep students focussed. It doesn't use a lot of bandwidth, either.
Just my $0.02
Think of welfare not as something which keeps "bludgers in booze money", but something that stops them from having to break into your house and kill you in order to survive.
:-)
Cut welfare, and that's what you get. Take away the safety net, and see what the cost to society is when tens of thousands of people start hitting rock bottom. Violent crime increases. You either pay for people to not do anything, or you pay for the consequences.
Unemployment is inevitable. We're been running at about 10% for as long as I can remember, most of that in youth unemployment. Given that we need less and less people every year to do the same work (all that 'efficiency' we keep saying is a good thing) the situation is only going to get worse. Oh, look what we're doing to Centrelink! That'll help!
All the stupid government make-work programs in the world aren't going to help this, either. We should make education free again, so they can go back into the universities where they belong
Seriously though, spending 10 years trying to get an Arts degree (ie: surfing and drinking beer on Austudy) is better than breaking and entering because there's no jobs, no money for this weeks' rent - or food - and no future.
As for your cretinous attitude to the Arts... it's on par with your misunderstanding of politics.
Sadly, the Australian government DOES listen to Microsoft. Every time Bill Gates comes out here, politicians fall over themselves to get photographed with him, our sad little prime-minister included.
It might also have something to do with MS's heavy involvement with the media over here. NineMSN particularly. (Think MSNBC) Politicians are always nice to people with significant media influence.
And partly it's the tired old "He's the richest man in the world, so he must be right."
Not that Microsoft doesn't ALREADY have a significant corporate presence here. They may not do much R&D, but their Sales and Certification departments are pretty big.
[I posted this in response to another comment, but I think it's relevant here too. :-)]
While I agree that the government here - on IT issues especially - couldn't find it's own arse if it used both hands and a map, I think Microsoft's suggestions are uniformly self-serving and would actually make the situation worse.
Our vaunted 'deregulation' has turned Telecom from a slothful but competent beauracracy into an agency that gouges more money from Australians than can be imagined, (biggest profits in australian corporate history) while sacking most of the technicians who actually run the system. Remember; they only started charging that $0.17/Mb in 1995 once they'd wrested control of AARNet from the vice-chancellors committee. As for the situation in the bush...
Frankly, what this nation needs is not more corporate investment and control, but:
* Nationalization of the telecom infrastructure (yes, you heard me) and deregulation only of services and support. Same for electricity, water, and all basic infrastructure; which is frankly why governments are supposed to exist.
* Significant government investment in education and basic research. Again.
* Sane and sensible policy on issues like digital signatures, privacy, and datacasting. (Die! Alston, Die!)
* Use of government clout to introduce new technology once past the incubator stage, which is where we have traditionally fallen down.
Of course, none of that is going to happen. I despair.
While I agree that the government here - on IT issues especially - couldn't find it's own arse if it used both hands and a map, I think Microsoft's suggestions are uniformly self-serving and would actually make the situation worse.
Our vaunted 'deregulation' has turned Telecom from a slothful but competent beauracracy into an agency that gouges more money from Australians than can be imagined, (biggest profits in australian corporate history) while sacking most of the technicians who actually run the system. Remember; they only started charging that $0.17/Mb in 1995 once they'd wrested control of AARNet from the vice-chancellors committee. As for the situation in the bush...
Frankly, what this nation needs is not more corporate investment and control, but:
* Nationalization of the telecom infrastructure (yes, you heard me) and deregulation only of services and support. Same for electricity, water, and all basic infrastructure; which is frankly why governments are supposed to exist.
* Significant government investment in education and basic research. Again.
* Sane and sensible policy on issues like digital signatures, privacy, and datacasting. (Die! Alston, Die!)
* Use of government clout to introduce new technology once past the incubator stage, which is where we have traditionally fallen down.
Of course, none of that is going to happen. I despair.
Sadly, I have to agree that Academic R&D is going downhill. CSIRO badly needs more funding, and the situation with Universities is going critical. (Literally, with the 'academic bribery' allegations currently flying around like neutrons in a reactor)
I agree that a lot of good R&D gets done interally by companies, possibly even more in total (I myself would probably be considered such) but you state exactly the problem: It gets exported. Or quietly dropped.
At least with academic research, the tradition of publishing everything means that whatever reasearch is actually done adds to the total pool of knowledge, if only to show "this doesn't work". Too much corporate research never sees the light of day, especially the failures, which are often the most important.
I know this is a generalization, but too much corporate 'R&D' crosses the line to 'Applied', in my opinion. Exactly what the academics are being pushed towards.
Oh boy. This is NOT going down well.
.NET. Make of that what you will.
:-) Most of that 'piracy' happens on an individual level. Software swapping with friends, etc. Small and 'backyard' businesses run a lot of 'grey' software. Large companies tend to be utterly legit, though.
I really should read the whole 'Accellerating Innovation' paper before commenting on details, but I can give a first-glance perspective:
On Bandwidth:
Australia was significantly behind the US in terms of broadband until about six months ago. Cable and DSL have arrived here with a vengance! $60 a month will get you 400kbit/s DSL. Really. The technical implementation has been rock-solid: none of the problems the US has been having. There are still political issues with the 'duopoly' telco situation, but the ACCC (Australian Competition and Consumer Comission) should do a good job of sorting out the more draconian stupidities in their 'Acceptable Usage Agreements' if they try to enforce them.
(The ACCC is way cool. They're currently having a spat with the MPAA over DVD regional coding.)
Home users here are in the best position they've ever been in. The biggest problem is that Telstra keeps running out of cable modems.
Academic and Business users are, however, in a position of paying rather more for interet access than is really sane. Telstra and Optus C&W both charge about $0.17 per megabyte for inbound traffic, (yes, you heard right) something which deregulation should have put an end to. But Optus likes the money too. Things should change with the recent landing of the Southern Cross Network pacific fibre.
Microsoft wants more broadband in Australia because they've targeted us as one of the prime test markets for
As far as accessibility goes, anyone who wants net access can get it. It's in every library, government department, and medium to large business. ISP's and net cafe's abound. The only people who don't have access are those too lazy to get off their fat arse and walk down the street.
On Taxes:
Yes, Australia has fairly high income and corporate taxes. We have correspondingly higher 'socialist' programs like decent state-sponsored Education, Medical care and Welfare. Australians are, on the whole, happy with this situation. Similar to Canada, in many respects. Multinationals don't like high taxes, though.
On R&D:
There is a cultural difference in what R&D means. We still have a very strong Academic research tradition: most really innovative stuff in Oz gets done by the CSIRO (commonwealth science and industry research organisation) and the universities. I think this is fine. Corporations suck at real R&D. I think Microsoft are just complaining they can't twiddle their thumbs inventing things like "Microsoft Bob" and get massive tax rebates for it.
Besides, what most US companies would call R&D, we call 'getting shit done'. We're an emminently practical people in many ways.
Failure to create local centres of excellence:
Rubbish. See above. The main issue seem to be the existing centres of excellence contain people who work outside the corporate domain and are, therefore, fairly unpoachable. ie: No talent pool waiting for Microsoft to snaffle them all.
On Piracy:
Australia has one of the highest rates of 'Piracy' in the western world. And literacy too. We are significant users of information products, even if we can't quite afford it.
There are some historical reasons for that, mostly due to semi-monopoly locks on imports of information products (the mentioned "parallel import restrictions") like software and music that kept the price artificially high for a long time. These have been changed recently, and will change again, since anyone with a net connection and credit card essentially ignores it.
A skills shortage:
"Australia faces many specific hurdles to train and retain skilled people."
Duh!! You know why? Because of the famous 'brain drain'. US/UK companies tend to make our best and brightest offers they can't refuse. We'd have better people if Microsoft stopped poaching them all.
On Capital:
Australia doesn't have anywhere near the level of Venture Capitalism that exists in the states. (but who does?) Yes, this does hurt us. He has a point here. But, if microsoft want to fix it, all they need to is start investing themselves. A billion US dollars goes a long, long way here. Microsoft has no lack of access to capital. I think what they mean is; there's no easy pool of money for Microsoft to suck from.
Technology security:
I can't think of any Oz sites that have been DoS'd, or even defaced. (I'm sure there's been some. Nothing to the level of Yahoo or Ebay. Mostly a US phenomenon) All major banks in Australia now offer internet service, and I've never heard of any problems. What has the writer been smoking?
Summary:
Yes, as many people would expect, Microsoft are complaining about local factors that tend to favor individuals and small businesses over large multinationals like them. There are one or two decent points in the dross, specifically those about lack of telecom deregulation and capital availability.
The gist of the complaint seems to be "give us more money or we won't invade your nation".
Sounds fine to me.
Ahhh... sugar. Is there anything it can't do?
Fascinating response.
:-)
If you really do want to know what's next, it will be an XML OO language that looks something like this:
<package name="example">
<knows n="net.oldgods.thor.example.include"/>
<object name="foo" inherits="bar">
<implements n="sortable"/>
<private>
<var name="x" v="3"/>
<var name="tree><x><y/><z/></x& gt;</var>
</private>
<public>
<property name="content" get="tree" set="tree"/>
<method name="order" c="integer">
<return v="$x"/>
</method>
</public>
</object>
</package>
Or, if it doesn't, will be transformable to it by stylesheet
"Um, and make me blind? "
/dev/null. This was written in two weeks with absolutely no prior experience of J3D. I say again: Java3D rocks. It will rock even more so when they finish it :-)
h'Wha? Remember, I've done this. Transparency works. It's easier to show than explain, so I've put up some screenshots here:
http://www.ecn.net.au/~orinoco/ideas/grid3d.jpg
Complaints about visual bugs can go straight to
"How is navigating a hierarchy with "hidden" objects/tasks/whatever better than a linear list of tasks/objects/whatever if the screen has plenty of space to display the tasks in a linear manner"
It's not. I didn't say anything about hierarchies or linear anythings. Take a look at the shots. The system I built actually works more on a hyperlink principle.
For anyone interested, it failed for two reasons:
1. Nobody had the hardware to run it.
2. Users kept getting lost.
Pretty, though. Ain't it?
Whoops. I've been doing too much Java3D. :-) When I said, "20% transparent", I really meant "20% opaque". You want to be able to see through about five layers of objects before they get completely occluded.
And don't go overboard on the lighting. Diffuse is enough. Specular lights are just distracting, even if they look cool.
Having recently spend two weeks building an ultimately unsucessful 3D database visualization system, (Java3D rocks!) I can tell you it's a LOT harder than it sounds, a bit like speech recognition.
The main problem is that there is no consistent paradigm for 3D interaction... no equivalent of a desktop metaphor. Users find themselves lost in space. And such systems are hard to interact with properly with a 2D screen and mouse - the missing degree of freedom create a 'modal' system that cannot be intuitive.
But, if you have to do it, here are Orinoco's tips:
1: Make everything about 20% transparent. You can't work with half the environment hidden behind the other half.
2: View control is the key. Don't make the user have to spin and rotate. Let them pick objects of interest, and then move the camera to a good view of it.
3: Don't try to model a complete 3D environment. Instead, make it "2.5D", with the extra dimension used to express an intrisic scalar quantity rather than a spatial one. eg: A 2-D scatterplot, but each point is instead a bar who's height indicates something.
4: Create a 'groundplane'. Stick to stacking things above this.
5: A 'recursive boxes' scheme, with whole new scenes hidden inside pickable objects works well. (A folder metaphor, if you will)
Just consider the most effective 3D application yet - 3D modelling. Even with a perfect 1:1 correspondence between the visual representation and the underlying model, it still takes experts to manipulate the interface.
Frankly, I think 3D interfaces have to wait until we have cheap and available 3D input devices.
Finally, there's a lot of research that has been done (SIGGRAPH, to name a source) that you would be silly to ignore.
I'm sorry. I can't buy into Froomkin's rant that ICANN is Evil because it's not under the control of the US Constitution. Hello? It's the INTERNET. Whatever process they come up with will conflict with some country's Constitution or ideology. That's why they were made independant... so they could function seperately. Duh! Please don't harp on how they're organised. It makes you look like a redneck to us 'Non-US Internet Users'.
He's got some legitimate concerns about the current agenda of ICANN and corporate influence, but I'm all for new TLDs... make hundreds! Use the entire dictionary, in fact! (Better than picking some poxy subset.
Lastly, DNS is fleeting. It is unlikely to survive for more than five to ten years, now. It's incredibly replaceable, not only with parallel DNS systems run by muliple 'authorities', (would make DNS a little more complicated to admin, but with some simple tools...) or you can simply turn your network nameserver into a gateway to Whatever Comes Next(tm), without having to touch any old clients, as long as the names still map to something.
There are numerous possibilities for totally distributed non-unique locator systems. Go talk to the mathematicians about Simulated Annealing, and the cryptologists about identity certificates. Hell, this latest paper on vertex saturation might be quite useful, too.
Finally, we should learn the lesson of centralism. DNS was a centralized point of power. It has been corrupted. So ends the lesson.
I will do this by means of character play: Corporate Server: Hi there! Welcome to GlobalCorp! Modified Mozilla Client: Yeah, hi. S: Please auto-download your details now. C: I don't want to. S: But you have to! C: I really don't want to. S: Then you can't have the page! Nyah. C: Oh, all right then. Bill CLinton. 1400 pensylvania avenue. Female. Guatamala. Sheep. S: Beep. Thank you. You may now read the page. C: [aside] Suckah. What P3P does is allow the machine to do all this for you! Transparently even! With no more of those stupid on-line forms. Mmmmm. Technology==Good. ~ Orinoco
To those in the know, this news seems almost expected. All scientists have curiosity, often triggered by what it right in front of them for much of the time. Therefore, it is only natural that some Australian scientist[1] should be sufficiently intrigued by strange bubbles in his beer, to mount a research effort of this magnitude. Helped, of course, by his supportive peers.
This is simply another step along the road to the GUB. (Grand Unified Beer) Slowly, we assembling a body of theory that will let us unify the four beers (Carlton, Swan, Guinness, and XXXX) at high energies. At Board length, some brewers theorise that 'Superbubbles' will form, and if properly chilled, it condenses into 'Superbeer'. Alas, it may be necessary to explode a nearby star to obtain the required energy. Ah well. No-one said good beer was easy. Please brace yourselves...
After this, we plan to turn our attention to Barbeque technology. We're impressed by the lighting times of the American Liquid Oxygen/Briquette BBQ, but have found it difficult to cook with, mostly since the steaks and sausages reduce to a charchoal impurity in the melted mass of the plate. We feel meat should be slightly rarer. Currently, we're tinkering with a solar furnace BBQ which is capable of cooking snags eight at a time at a distance of 30 metres.
In the ultimate model, orbiting solar mirrors are used to call down fire from the sky[2] into either permanent-emplacement BBQ recievers, or a portable 'campfire' version with GPS targeting system. Not to be used indoors.
Viva Australia! Mate.
[1]We are often defined by our environment. That which surrounds us, we are best at. Some prime examples where Australia still leads the world are; Solar Energy (due to a sun that will, given time, kill you with sheer brightness); and in developing antitoxins to nature's amazing array of very poisonous things. (Almost all of which live here, often in your boots)
[2]Clarke's law: Any technology sufficiently advanced is indistinguishable from magic)
Einstien simply has to be included. Enough said.
:-)
Same for Turing.
Feynman is up there. Pauli too. Strangely, I'd also add Ghandi to the list, perpetrator of the ultimate social hack, someone who made the future happen by living it just as hard as he could.
There are so many. Descartes. Bose. Schrodinger. Planck. Euler. (or was he too early?) Mandelbrot. Tesla. Fermi. Escher. Godel.
More recently and closer to home, The Woz, John West of DEC, (see: Soul of a new Machine) Ritchie and Kernigan. David Braben. Tim Berners-Lee. Bill Joy. Jaron Lanier. And of course, Donald E Knuth.
John Postel. Lest we forget. The whole IETF and W3C, for that matter.
Srinivatha Ramanujan.
Damn, I hope I spelled that right.
Without a doubt the greatest mathematician ever. His story is one of brilliance and tragedy. From a single old mathematics textbook, he extrapolated the entire of western mathematics, and went a hundred times further. Superstring theory depends on some of his math. And there are still notebooks full of his work which haven't been fully understood. He died young. Too young. Such loss is inexpressible.
Another brilliant but tragic geek is Taniyama, of the Shimura-Taniyama conjecture (again, I hope I got that right) which was recently finally proved, and formed the keystone of Fermat's last theorem[1]. He committed suicide.
For my part, I'd take Newton off any list due to his very ungeeky behavior outside physics[2]. Shockley is another I waver on. He invented the transistor, but wasn't very nice about it.
There's too many. The problem is we all stand on each others shoulders, all the way back to Aristotle. I'm reminded of a french mathematician priest who's enduring contribution to math was his letter writing, corresponding with dozens of mathematicians like Fourier, passing ideas and problems from one to the next. I don't remember his name. As a rule, most people don't either. I suspect the greatest geek of the century is one who made their mark in the same quiet, unobvious way. Their ideas so natural and obvious after the fact that no-one even wondered where they came from.
I say, leave the number one spot empty. Let it stand unfilled because we'll never be able to give it to the person who deserves it most.
~ Orinoco
[1] It's basically a mapping system which joins together two huge areas of mathematics. The reason it's important is that impossible problems in one domain can now be 'ported' across to another, solved there, and ported back.
[2] He was a cruel and vindictive man, who went out of his way to destroy the reputations of rivals. Brilliant, yes, but a nasty piece of work. I'm not surprised that people sat in trees and dropped fruit on him.