You can distinguish neutrinos from the sun and neutrinos from a generated source by just looking at the source (SuperK is a pointing neutrino telescope - it can tell where they came from).
The difference between neutrinos from the PP chain in the sun and generated neutrinos is that we will KNOW all the attributes of the neutrinos we generate (i.e. antineutrino vs. neutrino, muon neutrino versus tau neutrino versus electron neutrino, etc.)
We don't know what the neutrinos from the Sun look like. Just guesses.
(Neutrinos/ antineutrinos do annihilate. They don't ensure it - it's just that it would very rarely happen. The particle densities here aren't large enough to ensure constant interactions).
OK, for a moment I felt like responding, and disproving every point made in this post (you know, even as a troll, you could try to post, i dunno, real facts or something like that...) but nah.
I respond to trolls. Actually, on this specific topic, I'm not sure all of them are trolls - they don't have the 'troll-ific' nature that they normally do.
I find it easier to organize my thoughts when someone gives me a starting point, so I tend to respond more than I 'original-post'.
Mods note that I know this is offtopic, that's why I struck the score +1 bonus myself.:)
(yah, my UID is low. Should be lower, too, but I thought registering was for weenies for the longest time.:) )
How does bnetd violate their copyright? They're emulating a server - it's just as justified as SAMBA is, or (as I've pointed out eight times already...) just as justified as Bleem is, as Bleemcast! is, as Connectix VGS is, or ANY emulator is.
Blizzard copyrighted their GAME. They did NOT copyright the protocol that their game uses to interact with Battle.Net, nor could they (at least, not without someone being able to do a clean-room reverse engineer).
God, suck it up, Blizzard. Blizzard didn't even come UP with the "Online Play" idea in the first place! Why the hell are they the only ones allowed to play?
Ignoring CD-KEYS is specifically allowed in the DMCA, in case you missed THAT summary linked to in THIS ARTICLE.
I do understand the legal grounds on which Blizzard is making their argument - that is, none. Bleem was upheld even under the DMCA, so look, Blizzard has no chance.
And maybe Blizzard should find a way to stop warez copies, rather than stopping a Battle.Net server emulator. After all, even without the bnetd server emulator, there're still many different ways to play multiplayer games.
Look, piracy's basically a constant in terms of cost: it goes up a little some years, down some others. Why? Because pirates are smart people - and fundamentally, there's no way to prevent piracy - it's digital, and you can mess with it. The only way you could get rid of piracy is for a game like EverQuest, where the 'game' is located somewhere else, and EVEN THEN, someone could STILL make an additional 'EverQuest' server if they emulated it clean-room.
If Blizzard wants to stop piracy, stop the pirates, not the people who are doing perfectly legal things. But why aren't they doing this? Because it's bad for companies (and inefficient) to go after individual people. Too bad. Individual people are breaking the law. Sue them, not people who are doing LEGAL things.
Re:"Freedom" of thievery? Indeed.
on
EFF Takes Bnetd Case
·
· Score: 5, Insightful
Oh, please. Do you people NOT understand that reverse engineering is legal, and completely allowed. Do you also not understand that a free capitalist society WITHOUT the ability to reverse engineer would be a complete disaster? You wouldn't have two chipmakers competing in the x86 market, you wouldn't have any open source software whatsoever, not to mention that many scientific projects would be down the drain as well!
Blizzard sells games - people BUY games. Bnetd is not challenging the sale of those games - it's allowing people to play the games as they were intended - on a Battle.net server. Blizzard cannot, should not, and absolutely will not ever be able to say "oh, and you can only play these games on OUR server at Battle.net, and if you try to play them anywhere else, that's illegal."
bnetd is creating a free alternative to a necessary extra piece of 'hardware' to use the game to its full value. This is IDENTICAL to the Bleem/Connectix case. Completely identical. In that case, they created a free implementation of a Sony PlayStation on a computer, and the courts agreed "well, yes, if people BUY the games, they can attempt to PLAY them on whatever they want." If you try to say that bnetd is wrong, then by extension WINE is wrong, Bleem is wrong, hell, VMware is wrong. But they're not. Why? Because reverse engineering is legal, and critical, for any healthy economy.
Blizzard makes money off of games. In fact, Battle.net earns them no value, no reward in your own terminology. God. Blizzard, and everyone else, are just plain stupid here - let bnetd run, and Blizzard makes MORE money, for crying out loud. How the heck can it be bad to offer people MORE ways to play a game you're selling?!?
That's not the point. What they're trying to say is that a program which is interacting with another program (bnetd interacting with Diablo II) does not have to respond to authentication methods of one of the program.
You can't say 'Diablo II is only allowed to interact with such-and-such programs' and more importantly, you REALLY wouldn't want to. What about input drivers, such as any of the Gravis programs which remap keys? That's a program "interoperating" with Diablo II. Does Blizzard really want to injoin those programs from working? So long as a program isn't written to specifically get around anti-piracy provisions of a second program (bnetd is NOT: it's written to provide a server. the lack of CD-KEY checking is necessitated by Blizzard. This is exactly the argument that Bleem won in their lawsuit vs. Sony).
Keep in mind that the TOA of Blizzard products isn't law, and bnetd is NOT violating that TOA - the person using the Blizzard product might be, but that's not bnetd's problem (again, Bleem/Connectix case...) In any case, somehow I doubt that that TOA is even valid for certain things (we REALLY REALLY need some judge to come and say "get real, this is crazy" on these idiotic software licenses)
Blizzard is REALLY STUPID to continue this case in the court. IANAL, but honestly, the case is just way too similar to the Bleem and Connectix VGS case to survive. Here, Blizzard/Viviendi have an uphill battle, whereas all the EFF has to do is start filing briefs with tons upon tons of references to previous case law.
Unfortunately, you need to realize that heat is much more difficult than you think. Heat is energy, plain and simple. And 10 MK is the value thrown around because that's the temperature such that on average, particles have enough energy to overcome the Coulomb barrier to fuse into the nucleus (actually, to tunnel through the Coulomb barrier).
So, thus, what you need to do, if you're not going to raise the ENTIRE thing to 10 MK, is find a way to give particles the same energy. Thus, you need to accelerate them somehow. So, there's your two different ways to generate fusion - create a steady state environment of 10 MK (the sun's approach) or find a way to accelerate a few particles enough to fuse as well.
The problem here is that accelerating particles is a LOT harder than heating them! Heating them you just have to throw energy at them. That's easy. OK, to get to 10 MK, you need to throw A LOT of energy at them, but still, there's no fundamental 'challenge'. Accelerating particles is a challenge - you're fighting against the second law of thermodynamics here.
So, WHATEVER you do, you need to find a way to generate a situation where you have particles with an average energy corresponding to 10 MK (I *think* it's E = (some constant)*kT but I'm probably wrong) and they're in a situation where they can slam into a deuterium particle before losing energy.
"Before losing energy" means you're probably going to be doing this in vacuum, and particle accelerators all basically use electromagnetism, so that's probably what you're going to try to do. It's highly unlikely that you'll ever find a material that has zero resistivity to your extremely high energy particles (it's just too easy to spallate other nuclei, collide, etc). Keep in mind that superconductors rely on the fact that electrons hop into the -lowest- energy state - keeping something with extremely high energy from transferring its energy to a lower energy object is really difficult (white dwarf stars do it, with densities beyond mortal comprehension).
Just one more point - ANYTHING that produces fusion via conventional methods is doing it via plasma physics. What the sonoluminescence guys are saying is that they're creating a 10 MK plasma. I unfortunately find that hard to believe. Give me neutrons, or give me death.:) Now, it might be that some exotic material can create collapsable bubbles that reach 10 MK, but even then, I doubt it will be helpful. Don't scorn plasma physicists - they can GENERATE fusion plenty good. It's just that it takes more energy to create it than you get out of it. I doubt that, in this case, you're getting any noticeable extra energy.
See other comment. Beauty of being an author is that your ideas live on even after you die, and can be reinterpreted later on. Someone may come up with a very convincing argument later on that Dick clearly stated that Deckard was a replicant.
I'm not arguing it wasn't clear - after I read the comment the third or fourth time, I decided that it sounded like I was placing Dick in the present tense, but it didn't sound THAT bad...
For instance, "In his work, LOTR, JRR Tolkien tells the story..." even though Tolkien's been dead for quite a while. Same idea.
(though again, I do agree that it kinda wasn't clear)
Sorry, the last copyright I saw on Do Androids Dream of Electric Sheep was 1972, not 1968 (probably a second printing) - the 10 years was the time that the book had been out before the movie was released. I was off by 4 years.:)
As for the past/present tense re: Dick, when discussing authors' works, I tend to use present tense - so "I doubt Dick himself is completely agnostic" means that I doubt Dick is completely agnostic in the writing of Do Androids Dream of Electric Sheep. The writing still exists in the present day, so to me, it's present tense. I agree it did sound like Dick was still alive, though.
I actually was going to reply to this myself explaining the present/past tense thing, but I figured someone else would point it out. Beauty of Slashdot.
I don't think you can nearly put the strength on "definitely not in the novel" - whether or not Deckard is a replicant is one of the big open questions in that book. Honestly, I thought it was fairly obvious Deckard was a replicant (it was hinted at quite often enough - Rachael, and then the not-included other police station was a strong hint IMHO anyway, along with Deckard's dispassionate approach, AND his only -slight- moral trepidations. It would've been much harder for me!). To me, Deckard definitely was a replicant, even from the book.
I again say that I don't see how it changes the ending. The book then becomes less about how humans deal with the unhuman and more about what IS human, and what is the 'moral superiority' that humans have over replicants?
If you want the "ambiguous and powerful" bit back, start then thinking about Deckard's place in the world around him. Why choose a replicant? Surely the replicant would find out that he is a replicant and do exactly what Deckard did, right? And the goal is to stop replicants. What if humans were *unable* to do the job Deckard did, because of exactly the same problem - because they couldn't justify killing the replicants in their mind either - it just wasn't right. So they figured that they could program a replicant who wouldn't have the same moral trepidations, because replicants don't. Unfortunately, as it turns out, they were wrong in that case as well.
Why would Deckard have difficulty choosing to save her? Because of the difficulty it presents inside himself. He doesn't know he's a replicant. Saving her, in some sense, strengthens the possibility that he's a replicant. Killing her returns him to blissful ignorance, but at her sacrifice. Note again, saving her means that he's admitting that what he's been told is wrong, and that there is no difference, morally, between replicants and humans (and then, of course, he has to start wondering just what IS human - after all, remember - they stress that is the only difference).
This really is the beauty of the original book, and it carries through to the movie as well, mostly, because the story is powerful EITHER WAY. Either decision is perfectly valid, although, as we've both proven, those who believe one answer will vehemently declare that it was obvious, and they can't see how anyone could have come to the other conclusion.
In any case, I don't think you should blame Ridley for leaning one way in this argument - I think everyone does. You obviously do. I obviously do. I'll bet Dick does as well (so, in an X-Filian sort of way, the truth may be out there).
Wait, I'm confused: Blade Runner is based on Philip K. Dick's "Do Androids Dream of Electric Sheep", right? Dick left it completely open as to whether or not Deckard was a replicant - or so he says. Honestly, it's fairly clear in the book that Deckard was a replicant.
Suggesting that somehow that demeans the meaning of the book is a little bit weak. Deckard realized that the replicants could be morally equivalent to humans, and therefore, by extension, so can he, so again, it's still a moral victory. It's not faulty programming, it's just simple logic on his part. It's an allusion to prejudice, really, and is essentially trying to ask, in a Biblical sense, whether or not those without sin are throwing the stones.
It really has nothing to do with Ridley's obsession, in this case: whether or not Deckard is a replicant is really one of the constant questions about the book, which has been out longer than the movie (10 years!) If it's Ridley's obsession, then it's thousands of thousands of other people's (including myself) obsessions as well, many of whom have never seen the movie.
The fact that Ridley chose sides in this isn't a big deal. I doubt that Dick himself is completely agnostic as to whether or not Deckard was a replicant. I don't think ANYONE can be truly agnostic on this argument - everyone who's read the book has an opinion.
You know, I find this somewhat funny - that they would portray THIS as what's wrong with the American health system. This isn't what's wrong at all - mainly because, well, this would never happen. Doctors aren't heartless, and in life-or-death situations, amazing things happen.
What's wrong with the American health system is what's never heard about - ordinary people. People who don't have health insurance who have real chronic health problems that limit the life they can live. I'm not saying that this is even an American problem - in many ways it's a world problem, but many other countries have worked around it.
Emergency medicine isn't the real issue - it's chronic medicine. That is, prescriptions - THAT'S what eat the real cost. In emergency medicine, amazing things happen and a lot of what goes on there isn't limited by HMOs. Yah. You'll find individual examples, yes, but it's not the problem that chronic medicine is.
Seniors really have it worst, but there are other people who get screwed over as well, because the cost of the prescription is utterly insane. Students, for example - most students are uninsured for a year or two in college simply because most health care plans don't cover students past 21 (I was lucky - mine covered me through 24, and I have a pathetically bad one through the University now).
Now here comes the question - people will say "oh, so sorry, americans have it so bad, paying for drugs while we scrounge for food" - like hell. The issue is that there's no damned reason these drugs have to cost as much as they do. It's not like they cost that much to make. Keep in mind that the majority of research done by drug making companies is to preserve their patent on drugs! This is insane! I mean, REALLY REALLY insane! The problem dogging the US health care system is the same problem which hurts health care world wide, and solving it would solve a lot of problems world wide, not just in the US. Many people in the US can't afford prescriptions. They sure as hell can't afford them OUTSIDE the US. Yes, if they lowered their prices we could afford them easily, but then tons of aid agencies would be able to help other countries get them as well. THIS fight, if it's fought right, wins out for everyone.
So, generic drugs don't get out to the WORLD (not the US, the WORLD) because drug companies are wasting the talents of good researchers to muck around with old drugs to make them repatentable.
Honestly, there's a simple, easy way to fix a lot of the health care problems in the US. Kill the damned ability of drug manufacturers to not develop anything new and still make money. Make them revert to what they are SUPPOSED to be doing if they're doing research: DOING RESEARCH. Suddenly, all the costs of drugs drops ridiculously, and the HMOs have money to burn on emergency medicine.
Hey. It's another thing which geeks like - yelling at the patent system. Someone needs to kill that dinosaur ridiculously fast. The idea that you can sit on your ass and make money of off one good idea the rest of your life is a total crock. If you're an inventor, invent. If you're a scientist, do research. If you're an engineer, engineer. God. Think of all the money these corporations would save if they just abandoned all of their infrastructure in protecting patents and actually concentrated on doing research.
The problem with this argument should be fairly obvious, though.
The main problem is the fact that they're not providing you with a browser: they're providing you with a browser that's impossible to remove. And people who claim that "well, it's nice to be able to enter stuff in the go window" are missing the point - There's NO reason that MS couldn't have made the OS able to accept a browser of any type as a file manager, provided it met some specifications (see GNOME's WM spec). Or use a different HTML renderer. But, no, they were scared of Netscape, and so they bundled IE in with Windows.
Think I'm crazy? What about this - what if Windows didn't allow you to change the default "Open" program for filetypes? How is this any different than what's going on now? The point is NOT that MS bundled these programs - look at Linux, for instance. If RedHat started bundling commercial programs with Linux, great - but the OS allows you to remove them.
So, I'm not saying "strip out the middleware". What I'm saying is "strip out the integration of the middleware into the OS" or "make the middleware removable". If MSN was set up in Windows to be the ONLY ISP, and any other ISP didn't have nearly the flexibility that MSN had under Windows (for no good reason other than Microsoft won't tell anyone what the APIs that MSN uses are), would that be fair? What the states and everyone else is saying is add everything you want, but DON'T BREAK THE LAW. MS has a monopoly. If you have a monopoly, you can't go around acting as if you don't - you have to act differently. Basically, you have to be very "nice" with your monopoly - not use it to bully around people or increase your business.
That's kindof what the antitrust laws are for. They acknowledge that monopolies sometimes occur, but that when they do, the company needs to somehow maintain the air of a competitive environment.
It's interesting that you chose "C" for raw speed, because I'd agree - for short development, and for resulting speed.
If you're doing something really dumb, like analyzing stupid data, C is kindof obvious. C++ works for huge apps and for large projects, but for me, C is just too simple. I don't mind re-implementing linked lists, or anything else, because it takes so little time if the program's only going to be five or six functions and maybe 2-300 lines of code.
My idea of reusing code is cut-and-paste from previous programs. Is this all inefficient? Mostly. But the programs I write from scratch are typically "I need something which does this" where "this" is something simple.
Granted, all of my code looks like crap. But I'll change that in time - for now, all I want is a program that works, and soon, and writing it in C++ is leisure time I don't have right now. when things die down, maybe.
(but most likely not... most of my programs are for an RTOS on an embedded system with no memory and no processor power. Sounds like fun, eh?)
He's only going about a few km or so. At that distance, you're
talking a distance-imposed latency of, oh, microsecond-ish.
(3 nanos/meter, 1000 meters/km, 3 microseconds/km).
Latency due to packet switching, "traffic jams", etc. are going
to far exceed that. After all, wireless travels at speed-of-light-ish,
which is actually faster than having wire run to your house.
I'm not saying there isn't additional latency due to the protocol -
it's just that I guarantee that the latency of wireless at 1 m
vs. the latency at wireless at 1 km, with sufficient signal
strength, is only gonna be about 3 micros longer. Note the
"sufficient signal strength". Check to see if you're dropping tons
of packets or something - your signal strength might be horrible.
You're dead right on the "easier on your eyes" thing. This was mentioned (I believe) in "Foundation and Chaos", the continuation of Asimov's series by some other authors (can't remember their names).
Paper has one thing beating the hell out of Ebooks: It's passive. It doesn't generate any light - you read the book by seeing different contrasts. That's not saying all Ebook like formats are active. LCDs are active (most... back to this in a moment) and CRTs are obviously active. However, paper is distinctly passive. Now, the thing is, so are Palm-like monochrome LCDs, and there are passive LCDs, but their contrast ratio is just flat out horrible. It's black on... what? Slightly greyish-brown? Horrible. (I will grant you that a Palm, with the Indiglo-like backlight on, is actually really readable - much more so than paper in the dark, at least.)
I have to diagree with you on the "Lots more portable" thing. It's not. Not for me, at least. I read fast - way too fast, I know, but that's how I read. If I go on a week-long trip, or even worse, a month long trip, I barely even bother bringing books along any more. It's not worth it - I finish them too quickly. I usually finish a 400-page book in a day, which means that I'll need to carry seven-eight books. That's why, on my last two month-long trips, I brought along my Palm with several books on my laptop to read. There's no WAY I would've had enough space for 10 books in my luggage with 30 days worth of clothing, supplies, and other things in there. But my Palm, and my laptop, both of which I already needed, take up MUCH less space.
There's a curve there, unfortunately. In small amounts, books are more portable (although a Palm is pretty portable), but electronic media is fixed-size regardless of capacity, whereas books scale roughly linear with capacity. That really starts to hurt when you need a lot of info.
Personally, I'm holding out for electronic ink, which will be passive, high contrast ratio, and massive storage capabilities in a confined space. (It would be trivial to add text highlighting, margin writing capabilities to this, too)
Sounds like heaven to me...
Re:We (probably) won't ever actually ACHIEVE AI
on
Arguing A.I.
·
· Score: 2
You raise an interesting point, one that hopefully I'll raise another interesting point to.
What causes the subjective perceptual experience is the fact that it can be subjective - the other two are meaningless. Is a mind without any ability to perceive a mind? Personally, I say yes - dreams are an inherent answer to that question, and though people may say that dreams only mirror experiences that we've already had, I tend to disagree with that. A mind completely separated from perception might be totally unlike anything we could imagine, but it would still be a mind in my opinion. Back in the day, I created a race for a story of mine in which one of the genders had a problem in the connections of the brain, so their entire existence was internal, and had no external connections. Is this still a person? I think so. You can easily imagine a person who has no connection to the outside world whatsoever. I wouldn't be so arrogant as to say that that person doesn't have a mind.
OK, so what causes something to be subjective? From where I stand, that's the ability to recognize a choice and to act upon it. The interesting thing is that programs don't do this, and people do - maybe. That's a big question, and one that I can't answer, but I'll get back to this in a moment. Programs don't do this - the output is known straight from the beginning. That is, there's no free will. Again, some would say humans don't have free will either, but we don't know. We DO know that computers don't - that is, the outcome of their actions is entirely given - until you introduce an external element that interacts in an unknown way.
So, interestingly enough, the question about whether or not computers can ever develop AI comes back to a question about humans. What is it about us that makes us non-deterministic? This really is the fundament of consciousness, after all - you'll never convince me that I'm deterministic - ever. Why? Because if you ask me a question, I KNOW I could've chosen the other. If you shove an instruction at a computer, there's no choice there. That's subjectivity - the subject - me - matters in the outcome. In an (ideal) computer's case, the subject - the computer - doesn't matter. When you add the "system" argument, the software+computer STILL doesn't matter. Until you add a non-deterministic element, everything will always be guaranteed.
You could weakly introduce random elements - say, using the current systime as a seed. But that's not truly random: it IS predictable given enough information re: the system. Any analog external sensing device might add the random element you need - but again, that means that AI needs perception, and I'm not convinced that natural intelligence does, which makes AI distinctly different than human intelligence. Still intriguing, but not necessarily hugely important. (Note, as I thought about this more: it doesn't necessarily mean that the AI needs perception, it just needs a non-deterministic element to its environment - i.e., a deterministic program - there's a name for this kind of ideal computer, and I forget what it is - can never be intelligent - or conscious)
If you hadn't guessed, my entire argument for saying that humans really have natural intelligence is based upon the fact that I think that our brains are truly random in nature - that is, there's some quantum-type element to the decision-making going on in our brain. Consciousness, and the soul, then, is just the perception of aggregated quantum choices which vastly affect a macroscopic scale. This same argument could be interpreted as saying that every particle in existence has a "soul", but you're reaching there - I think it's only when you aggregate a lot of quantum systems together that you get perception - that is, a whole lot of randomness with a whole lot of ability to affect future choices.
Maybe our brains will be shown to not possess any quantum nature - that is, all of the activity is happening on a macroscopic (cellular) level and would be entirely deterministic. I doubt that'll happen, I really do. It would make the universe too weird (and by my own version of Occam's Razor, the universe tends towards whatever makes it more boring).
What you're assuming is that everyone's ability to read code is equal. I disagree with the first poster that a bored sysadmin could actually find a security hole. Finding bugs is hard - really hard. Especially bizarre ones.
So, given that everyone has access to the code, if crackers can find a hole faster than everyone else. I think, honestly, the people who will find it are the people who are best at it - people who have worked in computer security a long time. They're not likely to be crackers, to be honest. So, for the most part, I think you're best off with an open-source model.
Plus, with the Microsoft system, it's not that you don't have access to the code - you just don't have access to the code's internals. You know what the functions are. It's pretty trivial to write a program to iterate through several thousand system calls looking for something to break. These are the kind of holes that are easy to find in closed-source models (they're also the kind that ARE found in closed source models.) Linux, however, probably is more vulnerable to wackier bugs like race conditions and so forth.
As for how to find it: you know, I wouldn't be surprised if the way a lot of already-exploited bugs were fixed was by someone actually obtaining the script-kiddie package. There, you HAVE to have the source code to understand how something's being screwed up. That said, I'm sure Microsoft can do the same, but, again, it's a lot easier with a helluva lot more people.
The thing here is, that's not the point. Just the fact that the code is available to you is the entire key - yes, it might not do you any good, and yes, you don't read through the code, but other people do. How many people does Microsoft have working for them? How many of them are working on security? (yes, all of them, now, in theory) And how many of them are security experts? You can't fix a hole you don't know exists. Therefore, finding the hole is the important thing, and this is DIFFICULT. It's like trying to find a bug that MIGHT exist in your program - even if you spend all of your time looking for it, you'll miss tens of other ones. With open source software, there are huge numbers of people working on things - literally, huge. And I'm sure a ton of security experts look at the Linux code - especially when a bug is found - to see if something else might be vulnerable.
The important thing here is that with Linux, you *know* when a vulnerability is found. Suppose it's a really really subtle one, and it takes MS a long time to fix it. Suppose the same in Linux. In the Linux case, you know about it, because someone else looked through the code, and if you're intelligent (and a sysadmin, for instance) you'll filter Bugtraq for anything that pertains to your system. Then, you get the advisory, and can disable whatever's causing the hole if there's no fix and the data's that sensitive.
What I'm trying to say is the strength of the open source security model is the fact that the people who are best at it (security experts - not all of them work for Microsoft) can find the bugs. With MS's security model, only MS can find the bugs, and finding bugs is hard. Very hard. I personally think they're crazy with their model. Even if they don't release the source code, explaining the bug can only help. Giving information to people smarter than you to help you with a problem can only help you - and there are a LOT of people who are very smart who don't work at Microsoft.
And Linux isn't really hard to secure. In fact, I think one of the distros ASKS you what services you want to allow in. If you only allow SSH in, you're not going to be vulnerable to much.
Interestingly enough, this feature of AGP is not really critical to increasing performance in games - in fact, it could be counterproductive to it.
The AGP GART (Graphics Address Remapping Table, I believe) maps "video card memory addresses" to "main memory addresses", i.e., it's to allow the graphics card to grab textures, etc. directly from main memory without going through the CPU.
Many motherboard manufacturers use this feature to provide on-board video without any dedicated memory so they don't have to include any additional memory for the graphics card.
Of course, since this blows so massively performance-wise, it's mostly abandoned now.
Is the GART actually useful for anything except extending the video card's onboard memory? I'm not really sure...
Actually, being picky, I believe that CGS units are SI - they just don't happen to be MKS.
Anything that is formed with an SI prefix and an SI unit is obviously SI, so CGS has to be SI. I also know that the weird CGS equivalents for certain MKS names (i.e. Gauss vs. tesla, dyne vs. newton) are SI as well, so there's no problem there either.
SI is really quite inclusive. Wasn't there something a while back that said that English units are now SI as well, so "English" is just a system choice? (including English units simply means defining a conversion: 2.54 cm/inch, for instance)
Those are Heaviside-Lorentz units (F=QQ/4pi r^2), at least according to any electromagnetic textbook I've ever seen.
You can make the speed of light 1 in any system you want, just by adjusting mu-naught (so long as it remains free), but that makes the Maxwell equations ucky.
A physicist will say 9.8 m/s^2, but an astrophysicist will say 2x10^33 grams. There's then the adage of "if you see things in a schematic like 10 cm x 50 cm, it's never been built".
Physics does typically use MKS - it's astrophysics that's on the CGS trip. There are reasons for it (What's the magnetic field of the earth? Bout a gauss - tesla are WAY too huge to be convenient except for certain wacky guys).
As for the reason for electrostatic units (e.s.u.), that's simple: they're not a holdover for anything, it's just that when you're not talking about things that are measured in a lab (like volts, amps) you might as well use convenient math, and for e.s.u., the Coloumb force is just qq/r^2. For astrophysics, that's the easiest way to do it.
There will always be a split in the physics community, though, between theorists and experimentalists. Keeping track of constants is a pain in theory - the math is hard enough already without shoving constants in front of everything, and the constants really confuse a lot of the underlying structure, as well. Therefore, theorists will always use natural units, with everything set to 1, basically (Heaviside-Lorentz units for electric charge: set epsilon-naught to 1, mu-naught to 1, so the speed of light is 1). Unfortunately, those numbers are ridiculously inconvenient for actually DOING any physics, since we don't live on a subatomic scale, so real physicists (er... experimentalists:) ) will always use MKS.
Slashdot Mirror
You can distinguish neutrinos from the sun and neutrinos from a generated source by just looking at the source (SuperK is a pointing neutrino telescope - it can tell where they came from).
The difference between neutrinos from the PP chain in the sun and generated neutrinos is that we will KNOW all the attributes of the neutrinos we generate (i.e. antineutrino vs. neutrino, muon neutrino versus tau neutrino versus electron neutrino, etc.)
We don't know what the neutrinos from the Sun look like. Just guesses.
(Neutrinos/ antineutrinos do annihilate. They don't ensure it - it's just that it would very rarely happen. The particle densities here aren't large enough to ensure constant interactions).
OK, for a moment I felt like responding, and disproving every point made in this post (you know, even as a troll, you could try to post, i dunno, real facts or something like that...) but nah.
I respond to trolls. Actually, on this specific topic, I'm not sure all of them are trolls - they don't have the 'troll-ific' nature that they normally do.
:)
:) )
I find it easier to organize my thoughts when someone gives me a starting point, so I tend to respond more than I 'original-post'.
Mods note that I know this is offtopic, that's why I struck the score +1 bonus myself.
(yah, my UID is low. Should be lower, too, but I thought registering was for weenies for the longest time.
How does bnetd violate their copyright? They're emulating a server - it's just as justified as SAMBA is, or (as I've pointed out eight times already...) just as justified as Bleem is, as Bleemcast! is, as Connectix VGS is, or ANY emulator is.
Blizzard copyrighted their GAME. They did NOT copyright the protocol that their game uses to interact with Battle.Net, nor could they (at least, not without someone being able to do a clean-room reverse engineer).
God, suck it up, Blizzard. Blizzard didn't even come UP with the "Online Play" idea in the first place! Why the hell are they the only ones allowed to play?
Ignoring CD-KEYS is specifically allowed in the DMCA, in case you missed THAT summary linked to in THIS ARTICLE.
I do understand the legal grounds on which Blizzard is making their argument - that is, none. Bleem was upheld even under the DMCA, so look, Blizzard has no chance.
And maybe Blizzard should find a way to stop warez copies, rather than stopping a Battle.Net server emulator. After all, even without the bnetd server emulator, there're still many different ways to play multiplayer games.
Look, piracy's basically a constant in terms of cost: it goes up a little some years, down some others. Why? Because pirates are smart people - and fundamentally, there's no way to prevent piracy - it's digital, and you can mess with it. The only way you could get rid of piracy is for a game like EverQuest, where the 'game' is located somewhere else, and EVEN THEN, someone could STILL make an additional 'EverQuest' server if they emulated it clean-room.
If Blizzard wants to stop piracy, stop the pirates, not the people who are doing perfectly legal things. But why aren't they doing this? Because it's bad for companies (and inefficient) to go after individual people. Too bad. Individual people are breaking the law. Sue them, not people who are doing LEGAL things.
Oh, please. Do you people NOT understand that reverse engineering is legal, and completely allowed. Do you also not understand that a free capitalist society WITHOUT the ability to reverse engineer would be a complete disaster? You wouldn't have two chipmakers competing in the x86 market, you wouldn't have any open source software whatsoever, not to mention that many scientific projects would be down the drain as well!
Blizzard sells games - people BUY games. Bnetd is not challenging the sale of those games - it's allowing people to play the games as they were intended - on a Battle.net server. Blizzard cannot, should not, and absolutely will not ever be able to say "oh, and you can only play these games on OUR server at Battle.net, and if you try to play them anywhere else, that's illegal."
bnetd is creating a free alternative to a necessary extra piece of 'hardware' to use the game to its full value. This is IDENTICAL to the Bleem/Connectix case. Completely identical. In that case, they created a free implementation of a Sony PlayStation on a computer, and the courts agreed "well, yes, if people BUY the games, they can attempt to PLAY them on whatever they want." If you try to say that bnetd is wrong, then by extension WINE is wrong, Bleem is wrong, hell, VMware is wrong. But they're not. Why? Because reverse engineering is legal, and critical, for any healthy economy.
Blizzard makes money off of games. In fact, Battle.net earns them no value, no reward in your own terminology. God. Blizzard, and everyone else, are just plain stupid here - let bnetd run, and Blizzard makes MORE money, for crying out loud. How the heck can it be bad to offer people MORE ways to play a game you're selling?!?
That's not the point. What they're trying to say is that a program which is interacting with another program (bnetd interacting with Diablo II) does not have to respond to authentication methods of one of the program.
You can't say 'Diablo II is only allowed to interact with such-and-such programs' and more importantly, you REALLY wouldn't want to. What about input drivers, such as any of the Gravis programs which remap keys? That's a program "interoperating" with Diablo II. Does Blizzard really want to injoin those programs from working?
So long as a program isn't written to specifically get around anti-piracy provisions of a second program (bnetd is NOT: it's written to provide a server. the lack of CD-KEY checking is necessitated by Blizzard. This is exactly the argument that Bleem won in their lawsuit vs. Sony).
Keep in mind that the TOA of Blizzard products isn't law, and bnetd is NOT violating that TOA - the person using the Blizzard product might be, but that's not bnetd's problem (again, Bleem/Connectix case...) In any case, somehow I doubt that that TOA is even valid for certain things (we REALLY REALLY need some judge to come and say "get real, this is crazy" on these idiotic software licenses)
Blizzard is REALLY STUPID to continue this case in the court. IANAL, but honestly, the case is just way too similar to the Bleem and Connectix VGS case to survive. Here, Blizzard/Viviendi have an uphill battle, whereas all the EFF has to do is start filing briefs with tons upon tons of references to previous case law.
Unfortunately, you need to realize that heat is much more difficult than you think. Heat is energy, plain and simple. And 10 MK is the value thrown around because that's the temperature such that on average, particles have enough energy to overcome the Coulomb barrier to fuse into the nucleus (actually, to tunnel through the Coulomb barrier).
:) Now, it might be that some exotic material can create collapsable bubbles that reach 10 MK, but even then, I doubt it will be helpful. Don't scorn plasma physicists - they can GENERATE fusion plenty good. It's just that it takes more energy to create it than you get out of it. I doubt that, in this case, you're getting any noticeable extra energy.
So, thus, what you need to do, if you're not going to raise the ENTIRE thing to 10 MK, is find a way to give particles the same energy. Thus, you need to accelerate them somehow. So, there's your two different ways to generate fusion - create a steady state environment of 10 MK (the sun's approach) or find a way to accelerate a few particles enough to fuse as well.
The problem here is that accelerating particles is a LOT harder than heating them! Heating them you just have to throw energy at them. That's easy. OK, to get to 10 MK, you need to throw A LOT of energy at them, but still, there's no fundamental 'challenge'. Accelerating particles is a challenge - you're fighting against the second law of thermodynamics here.
So, WHATEVER you do, you need to find a way to generate a situation where you have particles with an average energy corresponding to 10 MK (I *think* it's E = (some constant)*kT but I'm probably wrong) and they're in a situation where they can slam into a deuterium particle before losing energy.
"Before losing energy" means you're probably going to be doing this in vacuum, and particle accelerators all basically use electromagnetism, so that's probably what you're going to try to do. It's highly unlikely that you'll ever find a material that has zero resistivity to your extremely high energy particles (it's just too easy to spallate other nuclei, collide, etc). Keep in mind that superconductors rely on the fact that electrons hop into the -lowest- energy state - keeping something with extremely high energy from transferring its energy to a lower energy object is really difficult (white dwarf stars do it, with densities beyond mortal comprehension).
Just one more point - ANYTHING that produces fusion via conventional methods is doing it via plasma physics. What the sonoluminescence guys are saying is that they're creating a 10 MK plasma. I unfortunately find that hard to believe. Give me neutrons, or give me death.
See other comment. Beauty of being an author is that your ideas live on even after you die, and can be reinterpreted later on. Someone may come up with a very convincing argument later on that Dick clearly stated that Deckard was a replicant.
I'm not arguing it wasn't clear - after I read the comment the third or fourth time, I decided that it sounded like I was placing Dick in the present tense, but it didn't sound THAT bad...
For instance,
"In his work, LOTR, JRR Tolkien tells the story..." even though Tolkien's been dead for quite a while. Same idea.
(though again, I do agree that it kinda wasn't clear)
Sorry, the last copyright I saw on Do Androids Dream of Electric Sheep was 1972, not 1968 (probably a second printing) - the 10 years was the time that the book had been out before the movie was released. I was off by 4 years. :)
As for the past/present tense re: Dick, when discussing authors' works, I tend to use present tense - so "I doubt Dick himself is completely agnostic" means that I doubt Dick is completely agnostic in the writing of Do Androids Dream of Electric Sheep. The writing still exists in the present day, so to me, it's present tense. I agree it did sound like Dick was still alive, though.
I actually was going to reply to this myself explaining the present/past tense thing, but I figured someone else would point it out. Beauty of Slashdot.
OK, so see my other reply, but...
I don't think you can nearly put the strength on "definitely not in the novel" - whether or not Deckard is a replicant is one of the big open questions in that book. Honestly, I thought it was fairly obvious Deckard was a replicant (it was hinted at quite often enough - Rachael, and then the not-included other police station was a strong hint IMHO anyway, along with Deckard's dispassionate approach, AND his only -slight- moral trepidations. It would've been much harder for me!). To me, Deckard definitely was a replicant, even from the book.
I again say that I don't see how it changes the ending. The book then becomes less about how humans deal with the unhuman and more about what IS human, and what is the 'moral superiority' that humans have over replicants?
If you want the "ambiguous and powerful" bit back, start then thinking about Deckard's place in the world around him. Why choose a replicant? Surely the replicant would find out that he is a replicant and do exactly what Deckard did, right? And the goal is to stop replicants. What if humans were *unable* to do the job Deckard did, because of exactly the same problem - because they couldn't justify killing the replicants in their mind either - it just wasn't right. So they figured that they could program a replicant who wouldn't have the same moral trepidations, because replicants don't. Unfortunately, as it turns out, they were wrong in that case as well.
Why would Deckard have difficulty choosing to save her? Because of the difficulty it presents inside himself. He doesn't know he's a replicant. Saving her, in some sense, strengthens the possibility that he's a replicant. Killing her returns him to blissful ignorance, but at her sacrifice. Note again, saving her means that he's admitting that what he's been told is wrong, and that there is no difference, morally, between replicants and humans (and then, of course, he has to start wondering just what IS human - after all, remember - they stress that is the only difference).
This really is the beauty of the original book, and it carries through to the movie as well, mostly, because the story is powerful EITHER WAY. Either decision is perfectly valid, although, as we've both proven, those who believe one answer will vehemently declare that it was obvious, and they can't see how anyone could have come to the other conclusion.
In any case, I don't think you should blame Ridley for leaning one way in this argument - I think everyone does. You obviously do. I obviously do. I'll bet Dick does as well (so, in an X-Filian sort of way, the truth may be out there).
Wait, I'm confused: Blade Runner is based on Philip K. Dick's "Do Androids Dream of Electric Sheep", right? Dick left it completely open as to whether or not Deckard was a replicant - or so he says. Honestly, it's fairly clear in the book that Deckard was a replicant.
Suggesting that somehow that demeans the meaning of the book is a little bit weak. Deckard realized that the replicants could be morally equivalent to humans, and therefore, by extension, so can he, so again, it's still a moral victory. It's not faulty programming, it's just simple logic on his part. It's an allusion to prejudice, really, and is essentially trying to ask, in a Biblical sense, whether or not those without sin are throwing the stones.
It really has nothing to do with Ridley's obsession, in this case: whether or not Deckard is a replicant is really one of the constant questions about the book, which has been out longer than the movie (10 years!) If it's Ridley's obsession, then it's thousands of thousands of other people's (including myself) obsessions as well, many of whom have never seen the movie.
The fact that Ridley chose sides in this isn't a big deal. I doubt that Dick himself is completely agnostic as to whether or not Deckard was a replicant. I don't think ANYONE can be truly agnostic on this argument - everyone who's read the book has an opinion.
You know, I find this somewhat funny - that they would portray THIS as what's wrong with the American health system. This isn't what's wrong at all - mainly because, well, this would never happen. Doctors aren't heartless, and in life-or-death situations, amazing things happen.
What's wrong with the American health system is what's never heard about - ordinary people. People who don't have health insurance who have real chronic health problems that limit the life they can live. I'm not saying that this is even an American problem - in many ways it's a world problem, but many other countries have worked around it.
Emergency medicine isn't the real issue - it's chronic medicine. That is, prescriptions - THAT'S what eat the real cost. In emergency medicine, amazing things happen and a lot of what goes on there isn't limited by HMOs. Yah. You'll find individual examples, yes, but it's not the problem that chronic medicine is.
Seniors really have it worst, but there are other people who get screwed over as well, because the cost of the prescription is utterly insane. Students, for example - most students are uninsured for a year or two in college simply because most health care plans don't cover students past 21 (I was lucky - mine covered me through 24, and I have a pathetically bad one through the University now).
Now here comes the question - people will say "oh, so sorry, americans have it so bad, paying for drugs while we scrounge for food" - like hell. The issue is that there's no damned reason these drugs have to cost as much as they do. It's not like they cost that much to make. Keep in mind that the majority of research done by drug making companies is to preserve their patent on drugs! This is insane! I mean, REALLY REALLY insane! The problem dogging the US health care system is the same problem which hurts health care world wide, and solving it would solve a lot of problems world wide, not just in the US. Many people in the US can't afford prescriptions. They sure as hell can't afford them OUTSIDE the US. Yes, if they lowered their prices we could afford them easily, but then tons of aid agencies would be able to help other countries get them as well. THIS fight, if it's fought right, wins out for everyone.
So, generic drugs don't get out to the WORLD (not the US, the WORLD) because drug companies are wasting the talents of good researchers to muck around with old drugs to make them repatentable.
Honestly, there's a simple, easy way to fix a lot of the health care problems in the US. Kill the damned ability of drug manufacturers to not develop anything new and still make money. Make them revert to what they are SUPPOSED to be doing if they're doing research: DOING RESEARCH. Suddenly, all the costs of drugs drops ridiculously, and the HMOs have money to burn on emergency medicine.
Hey. It's another thing which geeks like - yelling at the patent system. Someone needs to kill that dinosaur ridiculously fast. The idea that you can sit on your ass and make money of off one good idea the rest of your life is a total crock. If you're an inventor, invent. If you're a scientist, do research. If you're an engineer, engineer. God. Think of all the money these corporations would save if they just abandoned all of their infrastructure in protecting patents and actually concentrated on doing research.
The problem with this argument should be fairly obvious, though.
The main problem is the fact that they're not providing you with a browser: they're providing you with a browser that's impossible to remove. And people who claim that "well, it's nice to be able to enter stuff in the go window" are missing the point - There's NO reason that MS couldn't have made the OS able to accept a browser of any type as a file manager, provided it met some specifications (see GNOME's WM spec). Or use a different HTML renderer. But, no, they were scared of Netscape, and so they bundled IE in with Windows.
Think I'm crazy? What about this - what if Windows didn't allow you to change the default "Open" program for filetypes? How is this any different than what's going on now? The point is NOT that MS bundled these programs - look at Linux, for instance. If RedHat started bundling commercial programs with Linux, great - but the OS allows you to remove them.
So, I'm not saying "strip out the middleware". What I'm saying is "strip out the integration of the middleware into the OS" or "make the middleware removable". If MSN was set up in Windows to be the ONLY ISP, and any other ISP didn't have nearly the flexibility that MSN had under Windows (for no good reason other than Microsoft won't tell anyone what the APIs that MSN uses are), would that be fair? What the states and everyone else is saying is add everything you want, but DON'T BREAK THE LAW. MS has a monopoly. If you have a monopoly, you can't go around acting as if you don't - you have to act differently. Basically, you have to be very "nice" with your monopoly - not use it to bully around people or increase your business.
That's kindof what the antitrust laws are for. They acknowledge that monopolies sometimes occur, but that when they do, the company needs to somehow maintain the air of a competitive environment.
It's interesting that you chose "C" for raw speed, because I'd agree - for short development, and for resulting speed.
If you're doing something really dumb, like analyzing stupid data, C is kindof obvious. C++ works for huge apps and for large projects, but for me, C is just too simple. I don't mind re-implementing linked lists, or anything else, because it takes so little time if the program's only going to be five or six functions and maybe 2-300 lines of code.
My idea of reusing code is cut-and-paste from previous programs. Is this all inefficient? Mostly. But the programs I write from scratch are typically "I need something which does this" where "this" is something simple.
Granted, all of my code looks like crap. But I'll change that in time - for now, all I want is a program that works, and soon, and writing it in C++ is leisure time I don't have right now. when things die down, maybe.
(but most likely not... most of my programs are for an RTOS on an embedded system with no memory and no processor power. Sounds like fun, eh?)
He's only going about a few km or so. At that distance, you're
talking a distance-imposed latency of, oh, microsecond-ish.
(3 nanos/meter, 1000 meters/km, 3 microseconds/km).
Latency due to packet switching, "traffic jams", etc. are going
to far exceed that. After all, wireless travels at speed-of-light-ish,
which is actually faster than having wire run to your house.
I'm not saying there isn't additional latency due to the protocol -
it's just that I guarantee that the latency of wireless at 1 m
vs. the latency at wireless at 1 km, with sufficient signal
strength, is only gonna be about 3 micros longer. Note the
"sufficient signal strength". Check to see if you're dropping tons
of packets or something - your signal strength might be horrible.
You're dead right on the "easier on your eyes" thing. This was mentioned (I believe) in "Foundation and Chaos", the continuation of Asimov's series by some other authors (can't remember their names).
Paper has one thing beating the hell out of Ebooks: It's passive. It doesn't generate any light - you read the book by seeing different contrasts. That's not saying all Ebook like formats are active. LCDs are active (most... back to this in a moment) and CRTs are obviously active. However, paper is distinctly passive. Now, the thing is, so are Palm-like monochrome LCDs, and there are passive LCDs, but their contrast ratio is just flat out horrible. It's black on... what? Slightly greyish-brown? Horrible. (I will grant you that a Palm, with the Indiglo-like backlight on, is actually really readable - much more so than paper in the dark, at least.)
I have to diagree with you on the "Lots more portable" thing. It's not. Not for me, at least. I read fast - way too fast, I know, but that's how I read. If I go on a week-long trip, or even worse, a month long trip, I barely even bother bringing books along any more. It's not worth it - I finish them too quickly. I usually finish a 400-page book in a day, which means that I'll need to carry seven-eight books. That's why, on my last two month-long trips, I brought along my Palm with several books on my laptop to read. There's no WAY I would've had enough space for 10 books in my luggage with 30 days worth of clothing, supplies, and other things in there. But my Palm, and my laptop, both of which I already needed, take up MUCH less space.
There's a curve there, unfortunately. In small amounts, books are more portable (although a Palm is pretty portable), but electronic media is fixed-size regardless of capacity, whereas books scale roughly linear with capacity. That really starts to hurt when you need a lot of info.
Personally, I'm holding out for electronic ink, which will be passive, high contrast ratio, and massive storage capabilities in a confined space. (It would be trivial to add text highlighting, margin writing capabilities to this, too)
Sounds like heaven to me...
You raise an interesting point, one that hopefully I'll raise another interesting point to.
What causes the subjective perceptual experience is the fact that it can be subjective - the other two are meaningless. Is a mind without any ability to perceive a mind? Personally, I say yes - dreams are an inherent answer to that question, and though people may say that dreams only mirror experiences that we've already had, I tend to disagree with that. A mind completely separated from perception might be totally unlike anything we could imagine, but it would still be a mind in my opinion. Back in the day, I created a race for a story of mine in which one of the genders had a problem in the connections of the brain, so their entire existence was internal, and had no external connections. Is this still a person? I think so. You can easily imagine a person who has no connection to the outside world whatsoever. I wouldn't be so arrogant as to say that that person doesn't have a mind.
OK, so what causes something to be subjective? From where I stand, that's the ability to recognize a choice and to act upon it. The interesting thing is that programs don't do this, and people do - maybe. That's a big question, and one that I can't answer, but I'll get back to this in a moment. Programs don't do this - the output is known straight from the beginning. That is, there's no free will. Again, some would say humans don't have free will either, but we don't know. We DO know that computers don't - that is, the outcome of their actions is entirely given - until you introduce an external element that interacts in an unknown way.
So, interestingly enough, the question about whether or not computers can ever develop AI comes back to a question about humans. What is it about us that makes us non-deterministic? This really is the fundament of consciousness, after all - you'll never convince me that I'm deterministic - ever. Why? Because if you ask me a question, I KNOW I could've chosen the other. If you shove an instruction at a computer, there's no choice there. That's subjectivity - the subject - me - matters in the outcome. In an (ideal) computer's case, the subject - the computer - doesn't matter. When you add the "system" argument, the software+computer STILL doesn't matter. Until you add a non-deterministic element, everything will always be guaranteed.
You could weakly introduce random elements - say, using the current systime as a seed. But that's not truly random: it IS predictable given enough information re: the system. Any analog external sensing device might add the random element you need - but again, that means that AI needs perception, and I'm not convinced that natural intelligence does, which makes AI distinctly different than human intelligence. Still intriguing, but not necessarily hugely important. (Note, as I thought about this more: it doesn't necessarily mean that the AI needs perception, it just needs a non-deterministic element to its environment - i.e., a deterministic program - there's a name for this kind of ideal computer, and I forget what it is - can never be intelligent - or conscious)
If you hadn't guessed, my entire argument for saying that humans really have natural intelligence is based upon the fact that I think that our brains are truly random in nature - that is, there's some quantum-type element to the decision-making going on in our brain. Consciousness, and the soul, then, is just the perception of aggregated quantum choices which vastly affect a macroscopic scale. This same argument could be interpreted as saying that every particle in existence has a "soul", but you're reaching there - I think it's only when you aggregate a lot of quantum systems together that you get perception - that is, a whole lot of randomness with a whole lot of ability to affect future choices.
Maybe our brains will be shown to not possess any quantum nature - that is, all of the activity is happening on a macroscopic (cellular) level and would be entirely deterministic. I doubt that'll happen, I really do. It would make the universe too weird (and by my own version of Occam's Razor, the universe tends towards whatever makes it more boring).
What you're assuming is that everyone's ability to read code is equal. I disagree with the first poster that a bored sysadmin could actually find a security hole. Finding bugs is hard - really hard. Especially bizarre ones.
So, given that everyone has access to the code, if crackers can find a hole faster than everyone else. I think, honestly, the people who will find it are the people who are best at it - people who have worked in computer security a long time. They're not likely to be crackers, to be honest. So, for the most part, I think you're best off with an open-source model.
Plus, with the Microsoft system, it's not that you don't have access to the code - you just don't have access to the code's internals. You know what the functions are. It's pretty trivial to write a program to iterate through several thousand system calls looking for something to break. These are the kind of holes that are easy to find in closed-source models (they're also the kind that ARE found in closed source models.) Linux, however, probably is more vulnerable to wackier bugs like race conditions and so forth.
As for how to find it: you know, I wouldn't be surprised if the way a lot of already-exploited bugs were fixed was by someone actually obtaining the script-kiddie package. There, you HAVE to have the source code to understand how something's being screwed up. That said, I'm sure Microsoft can do the same, but, again, it's a lot easier with a helluva lot more people.
The thing here is, that's not the point. Just the fact that the code is available to you is the entire key - yes, it might not do you any good, and yes, you don't read through the code, but other people do. How many people does Microsoft have working for them? How many of them are working on security? (yes, all of them, now, in theory) And how many of them are security experts? You can't fix a hole you don't know exists. Therefore, finding the hole is the important thing, and this is DIFFICULT. It's like trying to find a bug that MIGHT exist in your program - even if you spend all of your time looking for it, you'll miss tens of other ones. With open source software, there are huge numbers of people working on things - literally, huge. And I'm sure a ton of security experts look at the Linux code - especially when a bug is found - to see if something else might be vulnerable.
The important thing here is that with Linux, you *know* when a vulnerability is found. Suppose it's a really really subtle one, and it takes MS a long time to fix it. Suppose the same in Linux. In the Linux case, you know about it, because someone else looked through the code, and if you're intelligent (and a sysadmin, for instance) you'll filter Bugtraq for anything that pertains to your system. Then, you get the advisory, and can disable whatever's causing the hole if there's no fix and the data's that sensitive.
What I'm trying to say is the strength of the open source security model is the fact that the people who are best at it (security experts - not all of them work for Microsoft) can find the bugs. With MS's security model, only MS can find the bugs, and finding bugs is hard. Very hard. I personally think they're crazy with their model. Even if they don't release the source code, explaining the bug can only help. Giving information to people smarter than you to help you with a problem can only help you - and there are a LOT of people who are very smart who don't work at Microsoft.
And Linux isn't really hard to secure. In fact, I think one of the distros ASKS you what services you want to allow in. If you only allow SSH in, you're not going to be vulnerable to much.
Interestingly enough, this feature of AGP is not really critical to increasing performance in games - in fact, it could be counterproductive to it.
The AGP GART (Graphics Address Remapping Table, I believe) maps "video card memory addresses" to "main memory addresses", i.e., it's to allow the graphics card to grab textures, etc. directly from main memory without going through the CPU.
Many motherboard manufacturers use this feature to provide on-board video without any dedicated memory so they don't have to include any additional memory for the graphics card.
Of course, since this blows so massively performance-wise, it's mostly abandoned now.
Is the GART actually useful for anything except extending the video card's onboard memory? I'm not really sure...
Actually, being picky, I believe that CGS units are SI - they just don't happen to be MKS.
Anything that is formed with an SI prefix and an SI unit is obviously SI, so CGS has to be SI. I also know that the weird CGS equivalents for certain MKS names (i.e. Gauss vs. tesla, dyne vs. newton) are SI as well, so there's no problem there either.
SI is really quite inclusive. Wasn't there something a while back that said that English units are now SI as well, so "English" is just a system choice? (including English units simply means defining a conversion: 2.54 cm/inch, for instance)
Those are Heaviside-Lorentz units (F=QQ/4pi r^2), at least according to any electromagnetic textbook I've ever seen.
You can make the speed of light 1 in any system you want, just by adjusting mu-naught (so long as it remains free), but that makes the Maxwell equations ucky.
Actually, the standard set is:
:) ) will always use MKS.
MKS: physics
English: engineering
CGS: astrophysics
A physicist will say 9.8 m/s^2, but an astrophysicist will say 2x10^33 grams. There's then the adage of "if you see things in a schematic like 10 cm x 50 cm, it's never been built".
Physics does typically use MKS - it's astrophysics that's on the CGS trip. There are reasons for it (What's the magnetic field of the earth? Bout a gauss - tesla are WAY too huge to be convenient except for certain wacky guys).
As for the reason for electrostatic units (e.s.u.), that's simple: they're not a holdover for anything, it's just that when you're not talking about things that are measured in a lab (like volts, amps) you might as well use convenient math, and for e.s.u., the Coloumb force is just qq/r^2. For astrophysics, that's the easiest way to do it.
There will always be a split in the physics community, though, between theorists and experimentalists. Keeping track of constants is a pain in theory - the math is hard enough already without shoving constants in front of everything, and the constants really confuse a lot of the underlying structure, as well. Therefore, theorists will always use natural units, with everything set to 1, basically (Heaviside-Lorentz units for electric charge: set epsilon-naught to 1, mu-naught to 1, so the speed of light is 1). Unfortunately, those numbers are ridiculously inconvenient for actually DOING any physics, since we don't live on a subatomic scale, so real physicists (er... experimentalists
That link's broken. Can you repeat the question or email me directly?