There's a magazine in the UK called 'Private Eye'. One of the things it does is called 'Warballs', which is basically sending up anyone who references 11/9 inappropriately as a comment or justification for some action. Your comment is deserving of submission.
Do you expect the world to stop just because of an atrocity on your soil ? How arrogant can you get ? If you look long enough back in time, you'd probably be able to come up with a sufficiently-bad atrocity on every day of the year. You need perspective.
Yes, September 11th was "a bad thing". Yes, you should try to prevent it from happening again. Yes, you should mourn those who died. Yes, you should get on with your lives, and No, you should not try and associate anything you don't like with such an atrocity. Frankly, you cheapen it by doing so.
I don't particularly think this comment will get anywhere on Slashdot - it's a mostly-US board after all. It still needed saying.
Hmm. I'm not so sure that's the case any more. One of my colleagues lives in Scotland, in an obscure village about 25 miles out of Glasgow. He commutes to London to work 3 days a week (way to go, Easyjet:-) and works at home the remaining 2.
About a month ago, this little village got its' exchange upgraded, and he now has ADSL, making life a lot easier...
I think you have to get 200 people (down from 500) to say they want their exchange updated, and it goes on the list of exchanges to be upgraded...
My brother, living in Liverpool, had a cable modem before I did (I've got 2Mbps from NTL) - he's had it for so long, I can't remember when he got it... Maybe the people I know have just been lucky, but certainly it doesn't seem as bad as it's painted in the press...
Agree about the cellphone coverage though:-) Just 'cos A implies B, doesn't mean that B implies A...
"I'm pretty sure that this was possible using BeOS on my PII 450 and the onboard ATI Rage around 1998."
Now that's seriously impressive. I mean an incoming 720x576x25fps signal in YUV 422 is ~20Mbytes/sec (roughly similar in NTSC). So, you have video->memory of 20Mbytes, Memory->CPU of 20Mbytes, CPU transform work (for when dragging the mouse distorts the cube, or suddenly the surroundings are mirrored/fogged/whatever) on 20Mbytes/sec (!) back to memory at RGBA*1280*1024*75fps (384 Mbytes/sec), and then finally a transfer (possibly over AGP) of 1280x1024x3@75fps (280 Mbytes/sec).
That gives a grand total of PCI traffic of 444Mbytes/sec and then an AGP transfer of 280 MBytes/sec.
I find that highly unlikely on a PC (on the basis that it's impossible to transfer the data, let alone process it at that rate on the PC).
Unless of course, it was running at some pathetic low resolution (640x480 ? 320x240 ?), without the alpha, Z, and stencil buffers for shadows and mirroring effects, and at a low frame rate ?
Trivial on an O2, of course - just set the video-source and sink, let the graphics h/w know about the GL environment, and let it rip. About 1000 lines of source code, and bandwidth ain't a problem.
You realise you're comparing to an SGI that's maybe 6 years old, yes ?
Show me a PC from 6 years ago that could overlay video onto surfaces with special effects (warp, transform, etc.). Now rotate a cube with 6 of these video surfaces running in parallel (one per face) at any time.
"This hardware sucks because the program's crap" is almost never a good argument. Perhaps there's a mismatch. Perhaps the program is crap, but the hardware is cool. Perhaps.... (you get the idea)...
SGI's in general tend to be slow CPU's with massive internal bandwidth for throwing data around, and massively fast graphics for the day. If you're running a cpu-intensive program, then Intel is probably for you. If you want a graphics/media workstation, SGI is the way to go. Surprise, the Post/Film industry likes SGI's. Discreet Logic Flame/Inferno is still the dominant s/w, and it's head and shoulders above the rest.
a large organisation with immense political connections.
The ARU is a small (there are less than 100, as far as I know) organisation, that is very much accountable for its' actions. Every bullet shot has to be accounted for, as in: "I shot this bullet now because..." in an incident report. It's also civilian, not military.
If it wasn't accountable, I wouldn't have made the comment about the outcry over the "excessive" bullets used to kill people. See the other post in reply to my original comment for an explanation (I wasn't aware of the reason, myself).
You might call Britain a police state for other reasons, but to compare the SS to the ARU is simply untrue. In any event, I far prefer the idea of a small number of trained armed policemen to the idea of every man/woman in a police uniform having a gun at his/her hip....
I invite you to (any time you like) get on the wrong side of a "bobby".
Their job is only to stop/catch unarmed (or at least, without ranged-weapons) criminals anyway. A policeman with one of the standard-issue batons is significantly better armed than joe crook with a knife...
Any time there is a gun-toting idiot (briefly) around, the police just call in the armed-response unit (ARU). Much better-trained snipers who don't seem to care where they hit, so long as the bad-guy gets it. Similar to SWAT teams, I suppose.
Gun crime isn't much of an issue in the UK anyway. There's a pretty-persistent rumour of a shoot-to-kill policy amongst the armed police. Perhaps that's a contributory factor:-) I think I heard of someone being shot earlier in the year around where I live (NE London). The shooter was shot dead by the police ARU. There was some criticism over the fact that he was hit by more than five bullets, which seemed overkill...
[Note that I'm not at all opposed to the bad-guy being shot. If you play the game, you play by ALL the rules...]
Brazil was settled by Portuguese people. The culture of Brazil derives heavily from the culture of Portugal. So Brazil is "essentially" northern European.
A fine and particularly cogent argument, sir, with perhaps one small flaw...
Portugal is one of the most Southern European countries, North is usually denoted towards the top of the map, not the bottom.
Why don't we worry about, oh, loss of the skill of blacksmithing? Log cabin building?
Perhaps because the feeling is that blacksmithing and log-cabin building are replaced skills. If we didn't have any other methods for building houses or forging steel, you can bet that both those skills would be in very high demand. As they once were.
On the other hand, going into space is an expensive (we're still in the early part of the payback curve...) but potentially highly lucrative activity. Having the skills to capitalise on it when it *does* become less expensive requires a long (oh, say 50 years...) investment now, which the US is about halfway through.
[We don't need more nuclear weapons] Well, Bush apparently thinks we need them
Proof positive that they're not necessary then...
It would take at most a few years to re-build such a program. Japan and Germany, two nations without nuclear weapons, are expected to be able to produce nuclear weapons within less than a year from the point that they decide that they need them
Um, this isn't a valid comparison. Both those countries could probably produce a large aeroplane transport within a year as well. In both cases, they have the raw material and the knowledge.
A lot of what space is about is learning new skills, testing new ideas, and discovering what we need to know, and what we *will* need to know. You can't do that on the ground. It's that simple.
Anyway, at some point we're all going to want to get off this muddy ball of rock, because there'll have been a nuclear war, "grey goo" will be consuming the planet, some terrorist group will unleash a planet-wide plague, GM crops will turn into Triffids, or even that it's cheaper to mine asteroids than sift through the remaining resources on Earth... take your pick:-) At that point, knowing a little about space might be a good thing...
Nuclear weapons labs are there because the US continues to develop nuclear weapons actively and because nuclear weapons are a major part of US military planning.
Store/usr/dict/words in a database, or find a list of names on the web you can use
Get two words, separate them with a "." and call that the reference tag
Store the IP and the reference tag in the DB
Write the mailto: url using the reference tag as the name part of the email address.
The downside is 2 selects and an insert on a DB for every page, but most sites are database-driven now anyway, and those that aren't probably don't care about the delay...
As for getting the spammers not the harvesters, surely it's the spammers you pick up on... every time I've had spam, I've got the IP address of the sending server in the mail headers, it's just that I have to read it to adjudge it spam... They either use their own mailserver, or an open relay. Firewall it on port 25. End of story.
That's easy. I firewall them against all incoming traffic. No more spam from them, and frankly I don't care if the originator (even if innocent) suffers. If I happen to supply something they want, they can fix their damn IT systems before they get back online to me:-)
What they're doing is not illegal, but neither is what I'm doing...
I don't think I agree with anything you wrote:-) Just reading through, and this is longer than I intended, but what the hell...
Then why do I have an appendix? (Or slim body hairs?)
Because there is very little selective pressure to remove these low-cost (in evolutionary terms) additions to the body. This is assuming that you can get rid of X without affecting Y, which is a heck of an assumption - most of our body parts are created/regulated by the interaction over time of *lots* of different genetic codes, your overall genetic code is not a blueprint you can just erase part of... Besides, they're not useful *now*. They presumably were *once*, and they may yet be again. Not in our lifetime, I suspect:-) but possibly in the future...
Let's ignore the obvious rebuttal to your point (use of the brain's savant abilities is proportional; if we have a brain half the design, we might have half of the all-around intelligence) and focus on the evolutionary advantages of having unused brain tissue.
How do you *know* it's proportional ? It may be highly non-linear in nature. Intelligence could be an emergent property, as opposed to intrinsic. There could be a minimum (or maximum) neuron-quantity threshold for intelligence to occur, the decision-surface for relative intelligence could be as complex as a fractal plane. We don't know.
First off, we're able to survive brain damage much easier. Being able to be thwacked in the head and still bring food home--and maybe go out and hunt some more the next day--is an obvious evolutionary advantage.
I think you're overlooking the incredibly difficult process humans go through in childbirth. The non-assisted mortality rate (for both mother and child) is far higher than any other mammalian species on the planet. Primate females almost always give birth without excessive labour. Human females labour can last over several hours, although today the child is more likely to be induced or surgically delivered. Only 200 years ago, death in childbirth was commonplace for those who could not afford assistance.
In contrast, being hit on the head hard enough to significantly break the skull will pretty much cause damage whatever size brain you have. Since all the higher-order functionality is on the outside of the brain (grey matter), that's the area that would be damaged anyway. If you don't break the skull, you're likely to just get a bruise either way, so long as you don't make a habit of it...
Don't forget that (unless our ancestors were particularly keen on headbutting cliffs) this would be an effect on 1 person. The do-or-die childbirth thing is an issue for every human born. I suspect nature might come down on the side of the majority...
Secondly, it increases mating. Having a bigger brain means our heads are shaped different--in a more asthetically pleasing fashion. The face is a human's primary means of identification and emotional communication--a clearer face is an obvious evolutionary advantage, within the species.
Um. No. If we all had faces the size of pygmy monkeys, we'd probably have designs on our chests or backs, or some other method of recognition. Sexual preference is closely tied to genetic fitness, not the other way around.
Consider that healthy-but-pug-ugly A has a 85% chance of surviving to breeding-age (and hanging around afterwards for protection etc.) because he's got strong arms. Handsome bigheaded B has only a 50% chance of making it, but he looks really cool. Unfortunately for B, the numbers are against him. No matter how many doting females are queueing up (hah!), if he only has a 50% chance of making it, his genes (and those of the doting females, since they choose B) are far more likely to be swept down evolution's sewer. The corollary is that the female
there was even an X-File about it.. some shack in the middle of nowhere with a T1.
Hmm. Maybe Alan's gone over to the dark side.... Sounds dangerously like an i-loo, or whatever that crap is called that Microsoft were pushing. Yeah, pun intended:-)
Funny perhaps, but an interesting point - if we did detect something from 1M years ago, why would they have come our way in the meantime ?
I mean, *light* has only just got here, and galactically speaking, we were pretty boring a million years ago (hell, in even inter-solar-system terms, we're pretty boring now!) I wouldn't get out of bed to travel a million light years to see if there's something here...
So, they may have colonised their entire sector/galaxy/galactic cluster using weirdo-science space travel; just because they didn't make it here yet, doesn't mean they didn't/couldn't...
Maybe in the US you can get a 1.5GHz machine for $200, over here in the UK, a minimal machine (cheapest 1.5GHz I could find flicking through Computer shopper) is 440 pounds ($660), including taxes. A PS2 is 149 pounds ($225).
You don't get SSE2 on Celerons or Athlons, as far as I know (you might get it on Celerons, but only the new Barton core from AMD does SSE2, and that's only on 3GHz+ machines, I think). My 1.5GHz machine above only has SSE/3dNow anyway.
According to (http://www.tommesani.com/MMXExamples.html), if I'm reading correctly, it takes 8 MMX instructions to do an 8-point (2x4-way) dot product. On the PS2, it would take 2 clocks on a single VU, or 1 clock if the algorithm was using both VU's... [No you wouldn't interleave dot-products across VU's, I'm talking about if you set both VU's running on the same problem, and divided the data by 2, half each]
Still pretty convincing, for 3-year-old technology...
When Sony bring out the PS3, I was wondering if they might even take this a stage further. If you were to put an FPGA into a console, you could actually have programmable hardware at your disposal... *That* would be cool:-))
The vector units on the PS2 are far and away more powerful than the graphics card on the XBox. They're fully general-purpose vector processors, not a graphics chip co-pro. I *do* know a little about this topic - I've written a VU simulator...
Go to the Sony PS2 demo section on the PS2 linux site, and look at the VU demos there - or at least read about them. There are examples of marionette models being manipulated in response to the user input (x,y,z,buttons for impulses, etc.) on the controller. The physics is correct. The entire program runs on VU1...
These aren't graphics pipelines with programmable filters (Cg, for example), they're general-purpose CPU's with float and integer registers, maths operations (obviously), local single-cycle RAM (for programs and data), dma channels, interrupts, the works.
I'll start to take note of the Xbox graphics when you can download programs to it, tell it to execute them (until an event happens), register inbound/outbound dma data-transfers so it doesn't run out of data, and then let the main CPU get on with doing its own stuff. What's that, you say ? It can't ? Oh well. Shame.
The PS2 was an experiment in a new computing architecture - one that almost cost it dearly, since it's a pain in the *rse to program if you don't adapt to its' strengths rather than force it to use your own. It's a dataflow architecture - you download programs to the CPUs, then stream the data (vertices, colours, textures, etc.) through the programs from RAM using DMA and onto the graphics rasteriser.
The thinking is that most transformation programs can be expressed relatively succinctly, and that there's always more data than program anyway, so your algorithm for the bubbling ripple effect is (say) 6k, with (say) 2M of vertex,texture,lookup, etc. You feed in the same (initial conditions) data every frame, and a time clock, letting the VU program calculate the vertex manipulations... Because the data is always going to be much larger than the code, the overhead in switching VU programs during a frame is negligible, so do it whenever you need to...
When you think of the problem that the PS2 engineers were trying to solve, the architecture is very neat & very elegant... The PS2 has (for its' time) enormous bandwidth between on-chip modules, local cache RAM on all the processors, a general-purpose MIPS chip to keep things ticking over (run this one, wait,... run this one, wait...), and the VU's to do the heavy lifting from frame to frame.
In fairness, the more recent PC cards are almost getting there - Cg2 will be better (it'll have loops!) I think the maximum program length is currently 128 instructions as well . Whoosh. But at least it's getting better.
To try and compare the two is laughable at best. Yes, they both produce similar games, but they do it very differently under the hood, and this thread was about using the VU's in a cluster to form a supercomputer, after all... (I think that's misguided, myself, even using BSP you'd be hard pressed to cope with the latencies involved, but that's another story!)
This, you see, is the point - they were using the VU's !!!
The PS2 has 2 vector units (VU's), each does a 4-way (xyzw) floating point vector dot product with another vector in a single clock cycle, with 4 cycle internal latency. The chips are pipelined, so you can just keep feeding any 4-cycle instructions into it (90-odd% are 4-cycle, matrix deconvolution will take it more:-) and you get the answers 4 clocks later...
That's 8x2 (16:-) float multiply ops per clock, at 300 MHz if you keep the pipeline going. The performance of a 700MHz Celeron doesn't even come close, it's roughly 1/8th the speed...
[This is the third time of posting - hopefully it'll beat the filters this time. This text is here entirely so I have to take more than 20 seconds to post this comment... ]
"this not a troll I swear The good news is that people are finally understanding that X sucks, X is ugly, slow, stupid, a big pile of hacks and rustry code."
Yes, it is a troll.
X is one of the primary reasons I like Linux (or any unix). I don't want a remote desktop. I want remote programs. I want to be able to ssh into any remote computer (including those I can't physically get access to) and run editors with the display pointing back to me. Not a desktop, just the editors. On a typical day I'll have programs (mainly terminals and editors, but the occasional graphics program) open from over a dozen machines, all happily cohabiting on my single desktop... This lets me work remotely - I can cut'n'paste between/etc/cshrc locally and/etc/cshrc remotely with ease. I like this. You can prise it out of my cold dead hands, and not before.
If it's ugly for you (I assume you mean aesthetically challenged, here), then get a new distro; you know, the ones with the anti-aliased rendered displays, and use a decent window manager. Frankly, if you're not prepared to put some effort in yourself, you deserve what you get.
It's not slow, at least not as far as I can tell, even my old matrox card (G450) can do several hundred 800x600 (typical game res.) blits/second, a semi-decent graphics card should do much better. The DRI really helped here, and decent drivers take advantage: if you're on a crappy graphics card, or one without decent support, change.
There has been work done (by the X team and others) to check how much faster it could be made by removing the (AF_UNIX not AF_INET) socket transport when you're running local. The result: The kernel unix socket code was as fast as anything the X team could do to transfer data around. X also uses shared memory (ie: zero-copy) to "transfer" images (pixmaps) from the client to the server when running locally.
(This is actually a quote from g4dget, but I agree wholeheartedly, so I'm including it)
Overall, the idea that network transparency is some sort of special feature that one pays a high price for is nonsense: all major desktop operating systems run in protected mode, and most GUI applications run in a different context from the window system. X11 simply has been designed that way from the ground up, while Windows and Macintosh have evolved there from "direct mode" graphics. Network transparency in X11 is not so much an issue of IPC or how it does graphics--it uses IPC like all desktop windowing systems--but in having well-defined network transparent support for features like window management and configuration information. It's lack of those features in Windows and OS X that means that Windows and OS X are not network transparent.
In practice, XFree86 is a damned efficient window system that, when it has comparable drivers for the graphics cards, beats OS X handily in terms of performance and memory usage, and usually even beats Windows.
Certainly stupid it's not. The concepts behind it haven't changed for over a decase, and have yet to be surpassed. It's true that the client/server model has changed over time, with far-more-capable framebuffers than X originally had to play with, but the X-server has evolved to cope with this - witness the various "extensions" that have become standardised...
As for "big FAT slow ass", TinyX (in the XFree86 source tree) takes a whopping 860k of space or so (depends on server-side pixmaps) when running on a zaurus. Whoosh. Almost a megabyte there. Whenever you see memory sizes in Linux, they invariably include the RAM in the graphics card (which is memory mapped so it can be used with shared memory) and the pixmaps that have been requested to be stored within server ram by clients. "FAT" it's not.
The take-home message is: Don't just complain. If it bothers you that much then get off your backside and do something about it - either do it yourself or cajole others into doing it for you, maybe even hire someone, or go use Windows, whichever makes you happier. I'd get more-informed before making any decisions though.
[I'll ignore the "big pile of hacks and rustry (sic) code." part of your post, after all, it is a troll.]
It's also noted in the article that they're moving the PS2 chip to 90nm later this year, which gives the capability of 4 (rather than 2) Mbytes of RAM on-board. The Sony plant is supposed to be 3 years (not 4) so I guess Xmas 2006...
Perhaps a PS2.5 ? Maybe with 2 emotion chips on them ?
What gives the PS2 so much power is the DSP-like vector units, (2 per emotion engine), although what gives the vector units a lot of their power is the fast direct (DMA) link to the graphics system (which presumably wouldn't exist on the 2nd emotion chip if the 1st was handling the display), so maybe adding another emotion engine would just confuse matters.
Certainly doubling the local RAM would help with texture space (at something of a premium compared to PC games...) but the PS2 is supposed to be a dataflow architecture which doesn't therefore need large texture space. Maybe it would help port games from the PC with less rewrite though.
Simon
Re:We should avoid using "content" to describe thi
on
Geocoding All Content
·
· Score: 4, Insightful
... and people wonder why the image of GNU-addicts is so tarnished.
For [insert deity here]'s sake!
Q: What are you providing as a content-provider for X?
A: The contents of X.
Enough said. There are many important battles to be fought against too-greedy IP and copyright holders. This isn't one of them.
But take the whole machine and drop it on a planet where the ambient temperature is high enough to keep the plasma hot. As the propellor extracts energy, more heat flows into the machine. What's wrong with it now?
It's late on a Saturday night, I'm drunk, and I may be wrong, but haven't you just given it an external energy source ? What's the conceptual difference between that and a car engine with an everlasting supply of petrol ? I thought one of the requirements for a PPM is to not depend on external energy sources.
I also don't see why cooler seawater has reduced entropy. I can see it having reduced energy (temperature being a good measure of energy) but why more-ordered-but-identical-energy collections of water-molecules should be colder (ie: less energetic!) than less-ordered-but-identical-energy molecules escapes me.
I'm ignoring the various frictions and viscosity issues, and I can't get my head around your proton/electron argument anyway (too drunk:-)))
You're confusing 'perpetual motion' with 'perpetual motion machine'. A machine has to do Work (definition: The transfer of energy from one physical system to another), whereas a body in a perfect orbit does not.
The only Work the earth does is when it hits "stuff" in its' orbital path. Every time the earth hits a single atom in its' path, it slows down slightly, and spirals slightly farther in towards the sun. OTOH, it doesn't move too much closer because the earth has a very high mass, and the atom doesn't... There is also "stuff" hitting us from the 'back' side as well, which speeds up the earth slightly in the same fashion.
Slashdot Mirror
There's a magazine in the UK called 'Private Eye'. One of the things it does is called 'Warballs', which is basically sending up anyone who references 11/9 inappropriately as a comment or justification for some action. Your comment is deserving of submission.
Do you expect the world to stop just because of an atrocity on your soil ? How arrogant can you get ? If you look long enough back in time, you'd probably be able to come up with a sufficiently-bad atrocity on every day of the year. You need perspective.
Yes, September 11th was "a bad thing". Yes, you should try to prevent it from happening again. Yes, you should mourn those who died. Yes, you should get on with your lives, and No, you should not try and associate anything you don't like with such an atrocity. Frankly, you cheapen it by doing so.
I don't particularly think this comment will get anywhere on Slashdot - it's a mostly-US board after all. It still needed saying.
Simon.
For [insert random deity]'s sake! All you need to do is click on the link (you did see that 'gcj' was a link, didn't you ? ) ....
Simon
Hmm. I'm not so sure that's the case any more. One of my colleagues lives in Scotland, in an obscure village about 25 miles out of Glasgow. He commutes to London to work 3 days a week (way to go, Easyjet :-) and works at home the remaining 2.
:-) Just 'cos A implies B, doesn't mean that B implies A...
About a month ago, this little village got its' exchange upgraded, and he now has ADSL, making life a lot easier...
I think you have to get 200 people (down from 500) to say they want their exchange updated, and it goes on the list of exchanges to be upgraded...
My brother, living in Liverpool, had a cable modem before I did (I've got 2Mbps from NTL) - he's had it for so long, I can't remember when he got it... Maybe the people I know have just been lucky, but certainly it doesn't seem as bad as it's painted in the press...
Agree about the cellphone coverage though
Simon.
Now that's seriously impressive. I mean an incoming 720x576x25fps signal in YUV 422 is ~20Mbytes/sec (roughly similar in NTSC). So, you have video->memory of 20Mbytes, Memory->CPU of 20Mbytes, CPU transform work (for when dragging the mouse distorts the cube, or suddenly the surroundings are mirrored/fogged/whatever) on 20Mbytes/sec (!) back to memory at RGBA*1280*1024*75fps (384 Mbytes/sec), and then finally a transfer (possibly over AGP) of 1280x1024x3@75fps (280 Mbytes/sec).
That gives a grand total of PCI traffic of 444Mbytes/sec and then an AGP transfer of 280 MBytes/sec.
I find that highly unlikely on a PC (on the basis that it's impossible to transfer the data, let alone process it at that rate on the PC).
Unless of course, it was running at some pathetic low resolution (640x480 ? 320x240 ?), without the alpha, Z, and stencil buffers for shadows and mirroring effects, and at a low frame rate ?
Trivial on an O2, of course - just set the video-source and sink, let the graphics h/w know about the GL environment, and let it rip. About 1000 lines of source code, and bandwidth ain't a problem.
Horses for courses...
Simon.
You realise you're comparing to an SGI that's maybe 6 years old, yes ?
.... (you get the idea)...
Show me a PC from 6 years ago that could overlay video onto surfaces with special effects (warp, transform, etc.). Now rotate a cube with 6 of these video surfaces running in parallel (one per face) at any time.
"This hardware sucks because the program's crap" is almost never a good argument. Perhaps there's a mismatch. Perhaps the program is crap, but the hardware is cool. Perhaps
SGI's in general tend to be slow CPU's with massive internal bandwidth for throwing data around, and massively fast graphics for the day. If you're running a cpu-intensive program, then Intel is probably for you. If you want a graphics/media workstation, SGI is the way to go. Surprise, the Post/Film industry likes SGI's. Discreet Logic Flame/Inferno is still the dominant s/w, and it's head and shoulders above the rest.
Simon.
If I remember my history rightly, the SS was:
The ARU is a small (there are less than 100, as far as I know) organisation, that is very much accountable for its' actions. Every bullet shot has to be accounted for, as in: "I shot this bullet now because
If it wasn't accountable, I wouldn't have made the comment about the outcry over the "excessive" bullets used to kill people. See the other post in reply to my original comment for an explanation (I wasn't aware of the reason, myself).
You might call Britain a police state for other reasons, but to compare the SS to the ARU is simply untrue. In any event, I far prefer the idea of a small number of trained armed policemen to the idea of every man/woman in a police uniform having a gun at his/her hip....
Simon.
I invite you to (any time you like) get on the wrong side of a "bobby".
:-) I think I heard of someone being shot earlier in the year around where I live (NE London). The shooter was shot dead by the police ARU. There was some criticism over the fact that he was hit by more than five bullets, which seemed overkill...
Their job is only to stop/catch unarmed (or at least, without ranged-weapons) criminals anyway. A policeman with one of the standard-issue batons is significantly better armed than joe crook with a knife...
Any time there is a gun-toting idiot (briefly) around, the police just call in the armed-response unit (ARU). Much better-trained snipers who don't seem to care where they hit, so long as the bad-guy gets it. Similar to SWAT teams, I suppose.
Gun crime isn't much of an issue in the UK anyway. There's a pretty-persistent rumour of a shoot-to-kill policy amongst the armed police. Perhaps that's a contributory factor
[Note that I'm not at all opposed to the bad-guy being shot. If you play the game, you play by ALL the rules...]
Simon.
Brazil was settled by Portuguese people. The culture of Brazil derives heavily from the culture of Portugal. So Brazil is "essentially" northern European.
A fine and particularly cogent argument, sir, with perhaps one small flaw...
Portugal is one of the most Southern European countries, North is usually denoted towards the top of the map, not the bottom.
Why don't we worry about, oh, loss of the skill of blacksmithing? Log cabin building?
:-) At that point, knowing a little about space might be a good thing...
Perhaps because the feeling is that blacksmithing and log-cabin building are replaced skills. If we didn't have any other methods for building houses or forging steel, you can bet that both those skills would be in very high demand. As they once were.
On the other hand, going into space is an expensive (we're still in the early part of the payback curve...) but potentially highly lucrative activity. Having the skills to capitalise on it when it *does* become less expensive requires a long (oh, say 50 years...) investment now, which the US is about halfway through.
[We don't need more nuclear weapons]
Well, Bush apparently thinks we need them
Proof positive that they're not necessary then...
It would take at most a few years to re-build such a program. Japan and Germany, two nations without nuclear weapons, are expected to be able to produce nuclear weapons within less than a year from the point that they decide that they need them
Um, this isn't a valid comparison. Both those countries could probably produce a large aeroplane transport within a year as well. In both cases, they have the raw material and the knowledge.
A lot of what space is about is learning new skills, testing new ideas, and discovering what we need to know, and what we *will* need to know. You can't do that on the ground. It's that simple.
Anyway, at some point we're all going to want to get off this muddy ball of rock, because there'll have been a nuclear war, "grey goo" will be consuming the planet, some terrorist group will unleash a planet-wide plague, GM crops will turn into Triffids, or even that it's cheaper to mine asteroids than sift through the remaining resources on Earth... take your pick
Nuclear weapons labs are there because the US continues to develop nuclear weapons actively and because nuclear weapons are a major part of US military planning.
This I agree with.
Simon.
The downside is 2 selects and an insert on a DB for every page, but most sites are database-driven now anyway, and those that aren't probably don't care about the delay...
As for getting the spammers not the harvesters, surely it's the spammers you pick up on
Simon.
Agreed it's not particularly newsworthy - been doing it for ages...
:-)
If it gets N more people to do it though, we might just make spammers lives that little bit harder
Simon
That's easy. I firewall them against all incoming traffic. No more spam from them, and frankly I don't care if the originator (even if innocent) suffers. If I happen to supply something they want, they can fix their damn IT systems before they get back online to me :-)
What they're doing is not illegal, but neither is what I'm doing...
Simon
Because there is very little selective pressure to remove these low-cost (in evolutionary terms) additions to the body. This is assuming that you can get rid of X without affecting Y, which is a heck of an assumption - most of our body parts are created/regulated by the interaction over time of *lots* of different genetic codes, your overall genetic code is not a blueprint you can just erase part of... Besides, they're not useful *now*. They presumably were *once*, and they may yet be again. Not in our lifetime, I suspect :-) but possibly in the future...
How do you *know* it's proportional ? It may be highly non-linear in nature. Intelligence could be an emergent property, as opposed to intrinsic. There could be a minimum (or maximum) neuron-quantity threshold for intelligence to occur, the decision-surface for relative intelligence could be as complex as a fractal plane. We don't know.
I think you're overlooking the incredibly difficult process humans go through in childbirth. The non-assisted mortality rate (for both mother and child) is far higher than any other mammalian species on the planet. Primate females almost always give birth without excessive labour. Human females labour can last over several hours, although today the child is more likely to be induced or surgically delivered. Only 200 years ago, death in childbirth was commonplace for those who could not afford assistance.
In contrast, being hit on the head hard enough to significantly break the skull will pretty much cause damage whatever size brain you have. Since all the higher-order functionality is on the outside of the brain (grey matter), that's the area that would be damaged anyway. If you don't break the skull, you're likely to just get a bruise either way, so long as you don't make a habit of it...
Don't forget that (unless our ancestors were particularly keen on headbutting cliffs) this would be an effect on 1 person. The do-or-die childbirth thing is an issue for every human born. I suspect nature might come down on the side of the majority...
Um. No. If we all had faces the size of pygmy monkeys, we'd probably have designs on our chests or backs, or some other method of recognition. Sexual preference is closely tied to genetic fitness, not the other way around.
Consider that healthy-but-pug-ugly A has a 85% chance of surviving to breeding-age (and hanging around afterwards for protection etc.) because he's got strong arms. Handsome bigheaded B has only a 50% chance of making it, but he looks really cool. Unfortunately for B, the numbers are against him. No matter how many doting females are queueing up (hah!), if he only has a 50% chance of making it, his genes (and those of the doting females, since they choose B) are far more likely to be swept down evolution's sewer. The corollary is that the female
Hmm. Maybe Alan's gone over to the dark side.... Sounds dangerously like an i-loo, or whatever that crap is called that Microsoft were pushing. Yeah, pun intended
Simon
Funny perhaps, but an interesting point - if we did detect something from 1M years ago, why would they have come our way in the meantime ?
...
I mean, *light* has only just got here, and galactically speaking, we were pretty boring a million years ago (hell, in even inter-solar-system terms, we're pretty boring now!) I wouldn't get out of bed to travel a million light years to see if there's something here
So, they may have colonised their entire sector/galaxy/galactic cluster using weirdo-science space travel; just because they didn't make it here yet, doesn't mean they didn't/couldn't...
Simon.
Maybe in the US you can get a 1.5GHz machine for $200, over here in the UK, a minimal machine (cheapest 1.5GHz I could find flicking through Computer shopper) is 440 pounds ($660), including taxes. A PS2 is 149 pounds ($225).
:-))
You don't get SSE2 on Celerons or Athlons, as far as I know (you might get it on Celerons, but only the new Barton core from AMD does SSE2, and that's only on 3GHz+ machines, I think). My 1.5GHz machine above only has SSE/3dNow anyway.
According to (http://www.tommesani.com/MMXExamples.html), if I'm reading correctly, it takes 8 MMX instructions to do an 8-point (2x4-way) dot product. On the PS2, it would take 2 clocks on a single VU, or 1 clock if the algorithm was using both VU's... [No you wouldn't interleave dot-products across VU's, I'm talking about if you set both VU's running on the same problem, and divided the data by 2, half each]
Still pretty convincing, for 3-year-old technology...
When Sony bring out the PS3, I was wondering if they might even take this a stage further. If you were to put an FPGA into a console, you could actually have programmable hardware at your disposal... *That* would be cool
Simon.
The vector units on the PS2 are far and away more powerful than the graphics card on the XBox. They're fully general-purpose vector processors, not a graphics chip co-pro. I *do* know a little about this topic - I've written a VU simulator...
... The PS2 has (for its' time) enormous bandwidth between on-chip modules, local cache RAM on all the processors, a general-purpose MIPS chip to keep things ticking over (run this one, wait,... run this one, wait...), and the VU's to do the heavy lifting from frame to frame.
Go to the Sony PS2 demo section on the PS2 linux site, and look at the VU demos there - or at least read about them. There are examples of marionette models being manipulated in response to the user input (x,y,z,buttons for impulses, etc.) on the controller. The physics is correct. The entire program runs on VU1...
These aren't graphics pipelines with programmable filters (Cg, for example), they're general-purpose CPU's with float and integer registers, maths operations (obviously), local single-cycle RAM (for programs and data), dma channels, interrupts, the works.
I'll start to take note of the Xbox graphics when you can download programs to it, tell it to execute them (until an event happens), register inbound/outbound dma data-transfers so it doesn't run out of data, and then let the main CPU get on with doing its own stuff. What's that, you say ? It can't ? Oh well. Shame.
The PS2 was an experiment in a new computing architecture - one that almost cost it dearly, since it's a pain in the *rse to program if you don't adapt to its' strengths rather than force it to use your own. It's a dataflow architecture - you download programs to the CPUs, then stream the data (vertices, colours, textures, etc.) through the programs from RAM using DMA and onto the graphics rasteriser.
The thinking is that most transformation programs can be expressed relatively succinctly, and that there's always more data than program anyway, so your algorithm for the bubbling ripple effect is (say) 6k, with (say) 2M of vertex,texture,lookup, etc. You feed in the same (initial conditions) data every frame, and a time clock, letting the VU program calculate the vertex manipulations... Because the data is always going to be much larger than the code, the overhead in switching VU programs during a frame is negligible, so do it whenever you need to...
When you think of the problem that the PS2 engineers were trying to solve, the architecture is very neat & very elegant
In fairness, the more recent PC cards are almost getting there - Cg2 will be better (it'll have loops!) I think the maximum program length is currently 128 instructions as well . Whoosh. But at least it's getting better.
To try and compare the two is laughable at best. Yes, they both produce similar games, but they do it very differently under the hood, and this thread was about using the VU's in a cluster to form a supercomputer, after all... (I think that's misguided, myself, even using BSP you'd be hard pressed to cope with the latencies involved, but that's another story!)
Simon
This, you see, is the point - they were using the VU's !!!
:-) and you get the answers 4 clocks later...
:-) float multiply ops per clock, at 300 MHz if you keep the pipeline going. The performance of a 700MHz Celeron doesn't even come close, it's roughly 1/8th the speed...
The PS2 has 2 vector units (VU's), each does a 4-way (xyzw) floating point vector dot product with another vector in a single clock cycle, with 4 cycle internal latency. The chips are pipelined, so you can just keep feeding any 4-cycle instructions into it (90-odd% are 4-cycle, matrix deconvolution will take it more
That's 8x2 (16
Simon
:-)
Simon.
[This is the third time of posting - hopefully it'll beat the filters this time. This text is here entirely so I have to take more than 20 seconds to post this comment... ]
Yes, it is a troll.
X is one of the primary reasons I like Linux (or any unix). I don't want a remote desktop. I want remote programs. I want to be able to ssh into any remote computer (including those I can't physically get access to) and run editors with the display pointing back to me. Not a desktop, just the editors. On a typical day I'll have programs (mainly terminals and editors, but the occasional graphics program) open from over a dozen machines, all happily cohabiting on my single desktop... This lets me work remotely - I can cut'n'paste between
If it's ugly for you (I assume you mean aesthetically challenged, here), then get a new distro; you know, the ones with the anti-aliased rendered displays, and use a decent window manager. Frankly, if you're not prepared to put some effort in yourself, you deserve what you get.
It's not slow, at least not as far as I can tell, even my old matrox card (G450) can do several hundred 800x600 (typical game res.) blits/second, a semi-decent graphics card should do much better. The DRI really helped here, and decent drivers take advantage: if you're on a crappy graphics card, or one without decent support, change.
There has been work done (by the X team and others) to check how much faster it could be made by removing the (AF_UNIX not AF_INET) socket transport when you're running local. The result: The kernel unix socket code was as fast as anything the X team could do to transfer data around. X also uses shared memory (ie: zero-copy) to "transfer" images (pixmaps) from the client to the server when running locally.
(This is actually a quote from g4dget, but I agree wholeheartedly, so I'm including it)
Overall, the idea that network transparency is some sort of special feature that one pays a high price for is nonsense: all major desktop operating systems run in protected mode, and most GUI applications run in a different context from the window system. X11 simply has been designed that way from the ground up, while Windows and Macintosh have evolved there from "direct mode" graphics. Network transparency in X11 is not so much an issue of IPC or how it does graphics--it uses IPC like all desktop windowing systems--but in having well-defined network transparent support for features like window management and configuration information. It's lack of those features in Windows and OS X that means that Windows and OS X are not network transparent.
In practice, XFree86 is a damned efficient window system that, when it has comparable drivers for the graphics cards, beats OS X handily in terms of performance and memory usage, and usually even beats Windows.
Certainly stupid it's not. The concepts behind it haven't changed for over a decase, and have yet to be surpassed. It's true that the client/server model has changed over time, with far-more-capable framebuffers than X originally had to play with, but the X-server has evolved to cope with this - witness the various "extensions" that have become standardised...
As for "big FAT slow ass", TinyX (in the XFree86 source tree) takes a whopping 860k of space or so (depends on server-side pixmaps) when running on a zaurus. Whoosh. Almost a megabyte there. Whenever you see memory sizes in Linux, they invariably include the RAM in the graphics card (which is memory mapped so it can be used with shared memory) and the pixmaps that have been requested to be stored within server ram by clients. "FAT" it's not.
The take-home message is: Don't just complain. If it bothers you that much then get off your backside and do something about it - either do it yourself or cajole others into doing it for you, maybe even hire someone, or go use Windows, whichever makes you happier. I'd get more-informed before making any decisions though.
[I'll ignore the "big pile of hacks and rustry (sic) code." part of your post, after all, it is a troll.]
Simon.
It's also noted in the article that they're moving the PS2 chip to 90nm later this year, which gives the capability of 4 (rather than 2) Mbytes of RAM on-board. The Sony plant is supposed to be 3 years (not 4) so I guess Xmas 2006...
Perhaps a PS2.5 ? Maybe with 2 emotion chips on them ?
What gives the PS2 so much power is the DSP-like vector units, (2 per emotion engine), although what gives the vector units a lot of their power is the fast direct (DMA) link to the graphics system (which presumably wouldn't exist on the 2nd emotion chip if the 1st was handling the display), so maybe adding another emotion engine would just confuse matters.
Certainly doubling the local RAM would help with texture space (at something of a premium compared to PC games...) but the PS2 is supposed to be a dataflow architecture which doesn't therefore need large texture space. Maybe it would help port games from the PC with less rewrite though.
Simon
... and people wonder why the image of GNU-addicts is so tarnished.
For [insert deity here]'s sake!
Q: What are you providing as a content-provider for X?
A: The contents of X.
Enough said. There are many important battles to be fought against too-greedy IP and copyright holders. This isn't one of them.
Simon.
It's late on a Saturday night, I'm drunk, and I may be wrong, but haven't you just given it an external energy source ? What's the conceptual difference between that and a car engine with an everlasting supply of petrol ? I thought one of the requirements for a PPM is to not depend on external energy sources.
I also don't see why cooler seawater has reduced entropy. I can see it having reduced energy (temperature being a good measure of energy) but why more-ordered-but-identical-energy collections of water-molecules should be colder (ie: less energetic!) than less-ordered-but-identical-energy molecules escapes me.
I'm ignoring the various frictions and viscosity issues, and I can't get my head around your proton/electron argument anyway (too drunk
Simon.
And until it slams into something, it's simply in perpetual motion, it's not a perpetual motion machine...
A machine must do Work (definition: The transfer of energy from one physical system to another).
Perpetual motion is easy. A perpetual motion machine is impossible.
Simon.
(Getting tired of pointing out that machines have to DO something)
You're confusing 'perpetual motion' with 'perpetual motion machine'. A machine has to do Work (definition: The transfer of energy from one physical system to another), whereas a body in a perfect orbit does not.
The only Work the earth does is when it hits "stuff" in its' orbital path. Every time the earth hits a single atom in its' path, it slows down slightly, and spirals slightly farther in towards the sun. OTOH, it doesn't move too much closer because the earth has a very high mass, and the atom doesn't... There is also "stuff" hitting us from the 'back' side as well, which speeds up the earth slightly in the same fashion.
Simon.