The argument against that is why should one person be punished more than another for committing the same crime?
That is actually very easy to answer.
When people commit crimes, we punish them. We do so for three reasons;
In the hope that being punished for doing something wrong this time will prevent the person from doing it again. (individual preventive)
In the hope that others, seeing that this person gets punished for doing something, avoids doing the same thing. (generally preventive)
Protect society against the actions of a criminal by physically preventing him/her from doing it. (only in case of jail)
It is fairly obvious that atleast the first of these are pretty nonexistant for reach people with constant-sum fines. Saying that going 70mph where only 50 is allowed gets punished by $150 fine is the same as saying that anyone for whom that sum is without consequence can calmly go 70 all the time. Thus, the punishment is not working.
More reasonable is a fixed percentage of your annual salary. The *reality* is that fining Joe Average $1000 is a significant deterrent for Joe, while to i.e. Gates it's completely irrelevant, it's not even interest on his fortune for the 5 minutes the cop hold him.)
Such a system does indeed exist atleast for some areas in Scandinavia. For example, in Finland if you speed, you get a fine based on your income. And for example in Norway, if you fail to file your tax-declarations timely, you're fined based on your income and wealth.
I download their drivers, direct from their homepage, and compile the module.
Somewhat later I start to wonder what license the thing really is under.
So I use modinfo as per the standard. When I do so, the license is listed as "GPL".
This is the *intended* result. They *deliberately* made the string so that that would be displayed.
I assume they mean what they say, and treat the module as if it is GPL.
I don't think they'd have a leg to stand on if they later wanted to try to punish me for breaking their license.
Everything I did was perfectly reasonable. I was not in the least careless. The fact that I'm confused about their intended license is a *sole* result of the fact that they themselves *deliberately* misled me.
There's something called estoppel. It basically says a single entity cannot contradict itself to the detriment of others. Thus, one single company cannot at the same time deliberately have their modules print "GPL" when queried over the standard interface as to the license, and at the same time try to have me punished for believing that the module is in fact GPL.
That's not how it works in the real world. In the real world;
Even extensive searches does *not* guarantee that you find all relevant patents.
The language is often so unclear that it takes years of wrangling in court to settle if something infringes or not, what makes you think a quick literature-search would settle it ?
If you did read the patent, but concluded that your invention doesn't infringe, then you'll owe triple damages if it does infringe, because it'll be claimed you infringed willingly.
There are so many patents issued that even trying to stay on top of new say software-patents is more than a full time job, so if you do, you'll have no time to actually invent anything.
Time from filing to issuing is typically a year or two, and can be much longer. The best possible search offers precisely no help at *ALL* avoiding infringement of a filed but not-yet issued patent.
For software-developers, the best thing to do about patents is to figth politically against them, and otherwise ignore them as best as possible. The *only* way to be certain you won't infringe some software-patent is never to release any software at all.
Sort of. But you have to look at the limits too, not only how "many" orders there are for buy and sell.
If "Last" was 7, and there are tons of people who want to sell at 7, but hardly any buyers who want to pay more than say 5, it does indeed indicate the stock is likely to drop.
If however, Last was 7, and there are *tons* of people who want to sell their stock at 9 while there are only a few people who want to buy at 8, this does not indicate a downward trend.
I guess my point is, look at the price, not only the order-depth.
par2. It's a program mostly used for posting big binaries on usenet, but it works perfectly for this purpose.
The basic idea is that you split the file into n+x pieces in such a way that any n pieces is sufficient to recreate the entire file.
So, for example, assuming the recording of grandpa is 100MB, you use par2 to split it into (for example) 6 25MB pieces in such a way that *ANY* 4 of those pieces can be reassembled to the original file.
Offcourse you could just make multiple complete copies, but that would take more space. the 6*25MB pieces take only 1.5 times as much place as the original 100MB, but you're still safe aslong as no more than 2 of those 6 pieces go lost or corrupted.
Sure. But randomized source-port only helps when BSD is the client, makes no difference at all when BSD is the server as is much (as in orders of magnitude) more common. (because most services run on, and must run on, well-published ports)
So, those 20 people who use bsd as a network *client* are somewhat less likely to have their tcp-connections successfully attacked as those who use predictable source-ports. (still not 48000 times safer as Theo writes, predictable does not typically equate "100% guessable with 1 try")
This "vulnerability" is kinda lame really. Previously, people who didn't think about it very much, assumed that since to reset a TCP-connection you need to guess the sequence-number, the chance of successfully doing so would be no higher than 1 in 2^32.
This "vulnerability" only points out that infact tcp-implementations will accept as valid any sequence-number between $CURRENT and $CURRENT+$WINDOW_SIZE.
So, instead of needing to try 2^32 times, you need "only" to try (2^32)/$WINDOW_SIZE times. Still fairly hard under typical conditions.
Window-size is however typically proportional to bandwith and inversely proportional to delay, so it'll be easiest to exploit on a tcp-connection that has high bandwith, and high ping-times. For example any connection that goes over satelite. (those of you that knee-jerk and think that high bandwith and high ping can't coexist should go reread first-year networking-curriculum.)
Sure. I didn't say I don't know how to restart X, only that it seems that doing so is nessecary.
Your method works, unless your distro puts "DontZap" in as an option in the config. I think that for modern distros, the "approved" method is probably logging in as root and doing "service dm restart", but there's literally dozens of methods that will work, yours too, for most distros.
You're rigth. X does not need to come down for a driver install.
Afterall, all the install does is put some files at the apropriate locations, and change a few lines in a config-file.
X *does* however need to be restarted before it'll take *advantage* of the new drivers. Atleast I know of no way to tell a running X-server to go re-read it's config and re-initialise drivers.
No. But it'll less often accidentally delete stuff you really want than you would yourself if you manually waded thorough 1000 spams a day, attempting to find the dozen or so legitime email in between, and also trying to avoid wasting *Too* much time sorting the spam away.
Not really. But the question is not if it helps the MPAA. The question is if it helps the "progress of science and the useful arts" to use language from the constitution.
For example, it's my opinion that reducing copyrigths to oh, say, 15 years, would massively help the progress of the useful arts. Yes, it'd screw over a few humongously rich people who're today sitting on the few copyrigthed items that are a) old, and b) worth something, but the purpose of copyrigth-law was never to benefit the MPAA, or anyone else. The purpose was, the only legal purpose by the constitution, to benefit the progress of science and the useful arts.
That's not the point. The point of that explanation is to show how preventing aimbots is impossible, even in theory.
Noone is seriously suggesting that a significant number of players would cheat in this manner, only that the existence of this cheat-possibility demonstrates the (ultimate) futility of anti-cheat technology.
In the real world, a cheat can be software-only aslong as the user has complete control of his computer. The only way to prevent this is to take away the users control of the computer.
The principal provlem is that your game communicates with the world around it (including the graphics-card and the mouse and the network) by way of sending requests to the OS, and have them responded to. Aslong as the user controls the OS, he can have them responded to the way he likes, and there's nothing you can do about it.
In a computer the user controls, the OS can say: "mouse moved to x,y", and your game has no way whatsoever of verifying if this really happened, or if the OS is lying to you.
I don't know if you're willing to give up control of your computer and your OS and hand the keys over to "content creators", just in order to be able to enjoy a cheat-free online gaming experience. Personally I consider that a too high price to pay by something like 3 orders of magnitude.
They guy with more chips definately has an advantage because he can afford more mistakes. How is your example any different from a regular poker game? You against everyone else. The pot is not going to go to just one person everytime one of them wins. Anyway....
It is different from a normal poker game because, not only are the three cooperating with eachother and sharing their pot (people could do this in normal poker), but they *also* has access to a perfect communication-channel where they can freely exchange information with no chance of it being discovered.
Whether you are able to see it or not, that *does* make a real difference. The principal difference is that they can always *know* for *sure* which of the three has the best hand, so who should fold, and who should bet.
The reason folding the two worse hands make a difference is because, though it doesn't influence their chanse of winning, or the payout when they do. It *decreases* their losses on the occasions when they loose.
Essentially, it is easier to guess correctly the answer to the question: "Do I have the best hand", when you get to see two of the other hands.
You're rigth about more chips being an advantage. That is why in poker-competitions you usually start of everyone with the same initial stack of chips.
Speed in use. KDE seems (subjectively, I admit) to spend longer than XP in starting up, but to respond more snappily once it is up. To me that's a plus, because I don't constantly reboot my computer.
Flexibility. For example, the new ability to easily write your own handlers for given mime-types is a goodsend. For example, you can now, easily, say that for files of type image/jpeg right-clicking the file should give the additional menu-entry "Show exif-information". Or that right-clicking a folder (inode/directory) shall give the option "backup to backupserver". Doing such things requires nothing more than that you are able to give a shell-script command that does the wanted stuff.
Protocol-handlers. It's all nice and good that the file-manager can handle http:// aswell as file://, but with XP that's about where it ends. KDE 3.2 handles a lot more. cameras:// fish:// (access a host over ssh), smb:// and so on. Writing new kioslaves are a bit more tricky, but the collection is growing at a nice rate nevertheless.
Automation. dcop makes it possible to control the entire gui from a script, and gives flexibility to configure kde so it responds the way *you* want it to. Want IM messages from someone received in kopete to automatically ask kweather to look up, and show you, weather-info for the place where your pal lives ? No problem.
Freedom. Okay, so that's really mostly a feature of the license, and not a feature of the software, I happen to love it anyway. YMMV
That's a start. By the way, I'm not saying that XP doesn't do anything well. Maybe it does. You'll have to ask someone else about that though.
Ok, so here's how that'd work. With a 32-bit architecture (data and adress-bus) not only are the adresses 32-bit, but you also, typically, get 32 bits of data back from a adress.
So, if you set the adress "0000 0000", (yes, I know, that adress is magical and "forbidden" on some architectures, I'm simplifying here!) you get back 32 bits of data, stored from "0000 0000" to "0000 0003", 32 bits, 4 bytes.
If you ask for the data at "0000 0001" you'll instead get the stuff from "0000 0001" to "0000 0004", notice the overlap with the last call.
What this means is that there are 4 different ways of getting at every byte in memory. If you really want to read byte 17, you can do so by asking for adress 17, 16, 15 or 14, and discard those bytes you don't need. (this works because on 32-bit architectures, when you ask for "14", you really get 14-15-16-17, or atleast can at no extra delay. (because the bus doesn't go any faster with just 8 bit on it instead of 32 bits)
An alternative would be for a architecture to only be word adressable. This would mean that if you asked for adress 0 you get the first word in memory, which is bytes 0-1-2-3. If you ask for adress 1 you get bytes 4-5-6-7, in general, if you ask for address X you get word number X, which is bytes number X*4, X*4+1, X*4+2, X*4+3.
A "drawback" of this architecture would be that if you really wanted adress 17, you'd *have* to do it by asking for word 4, which would return you bytes 16-17-18-19, and then discard the unwanted stuff. Your CPU could have an opcode that did this for you though, and besides, we're pretty much doing this anyway, because frequently aligned (as in the adress I ask for is divisible by the number of bytes in a word) access is frequently faster than non-aligned.
You're rigth that sometimes there's "address bits" that aren't really part of the adress, so the effective adress-space is smaller. For example, for a long time Linux used 1 bit to determine if the address was "user space" or "kernel space", which effectively limited a single user-space program from having a adress-space larger than 2^32.
Good bridge players will know 100%, and a good player having won the bid can, in the later stages of the game, will know which player is holding each remaining card, based on their play history.
But with cheating even a complete newbie will know 100%. Besides, even the best bridge-couple in the country can't know 100%, from the get-go, what cards their partner has. Yes, we use the (wussname-in-english?) 'bidding' phase to communicate a bit to eachothers, but this is a very low bandwith channel, typical information is on the level of "I've got 13-20 points", if you've got strong cards and can bid high, you can transmit a bit more info, perhaps, for example, "Clover is my strongest colour", "I have few squares"
Knowing 100% of what your parner has (and thus 100% of what your two opponents have, only not who of them has what) is a humongous advantage.
Online play is still possible offcourse. But you'll be more or less forced to either accept constantly loosing to the cheating 12 year olds (which gets old real fast!), or do as I do -- play online only against people I know and trust.
This doesn't help. It makes cheating more difficult, but in no way impossible. I could have a second computer analyse the vga-out from the first, and generate the apropriate mouse-commands to send to the first.
The first computer would have no way of knowing that the commands came from a second computer, and not from a mouse held by a human.
*even* if you could somehow cryptographically ensure the integrity of the mouse, this would *still* gain you nothing, because there's nothing preventing a sufficiently determined person from creating a motorized trackball under the mouse that rotates under command of the second computer and thus creates mouse-movements.
The problem is analogous to the one of "copy-protecting" music, but still leaving the music hearable to the buyer. It's simply not possible, for the fairly fundamental reason that if you can hear it, so can a mic. Similarily, aimbot-proofing a online shooter is not possible because if a human could move a mouse like *so*, then the aimbot could do the same using mechanics.
You could say the mouse should use biometrics to ensure you're the one holding it. But a) I'm thinking you see that this is getting ridicolous and b) that also wouldn't help, I'd simply hold the mouse, and have a piece of the table move automatically under it so all I need to do is hold my hand more or less still.
Actually, if you are a good enough of a poker player this won't matter.
Yes it will. You are being disingenious. I *said* that when I said "cheating" I really meant unfair advantage. Are you saying that getting three hands dealed, and getting to choose which one to play in poker is *not* an advantage ?
The fact that an advantage *can* theoretically sometimes be overcome by a much better player is irrelevant. That is like saying that doping is not cheating aslong as there exists someone who can beat you even when you are doped and they are not. That's nonsense.
Problem is, the problem *is* cheating. Not only that the people win a lot. People *do* care if the person who just grind them into a fine dust is really a super-player, or if it's the teenage-kid newbie from down the block with a +700% health-patch.
One of the points of training in a game, or sport, is to be able to see yourself getting better, to increase your reputation.
Who would bother spending a lot of time becoming the best possible CS-player if all that means is that you'll be able to compete successfully in a group consisting of the newbie from last week that upped his health to ten times the normal ? Would that still be a challenge and fun ? Even when you know that what you achieve with an extra half-year of training, the kid will compensate by multiplying his health not by 10, but by 25 ?
Sorry, I just don't believe in your thesis. That people only care about an even match, and *don't* care at all about fairness and cheating. Handicaps is something different, the principal difference being the openness. There's a difference between saying "Mr Wood, I challenge you to a game, I recognize you are much better than me, so I propose we add 30 strokes to your result, and see if I can beat that." and saying "Mr Wood, I can kick your butt any day of the week" (because I have a cheat installed, but I won't say that.)
You're rigth. The elements does not in any way proove MS are behind them, not even really point in that direction.
What they *do* point at, pretty consistently, is a less than reputable, less than serious company. Not that it's a huge surprise that SCO has dealings with consultants of similar moral fiber.
As long as a vendor knows that you will buy his shit no matter what he demands for it, no matter how late it is, and no matter how it performs, he has no incentive to charge less or deliver more.
As a customer, you have zero negotiation-room aslong as both you and the vendor knows that he's got you by the balls.
His ball-grip *only* looses at the point where you seriously start to consider other vendors. What would your company if the next version of Windows cost $500/processor ? Suck it up or change vendors ? $1000 ? $10.000 ? Point is, sooner or later there comes a point where you'll say "No. I won't pay that. Sorry."
Going to a convicted monopolist and saying the equivalent of "I'll buy your stuff at ANY price, but could you please sell it to me cheaper anyway ?" is a nonstarter.
These days, however, going to Micrsoft and saying "Look, upgrading to the newest Office for our 700 desktops is going to cost $X, we like Office, mostly, but the thing is, we also got an offer for something called OpenOffice, and it seems that even with retraining-costs it'll still only be $0.5X, and besides, the upgrades thereafter are free." has a tendency to get you offered 50-90% discounts on the MS-software.
This demonstrates perfectly *exactly* where the ball-grip has it's roots. It exists only when Microsoft feels sure you'll buy their stuff anyway.
Most computer-architectures are not bitwise-adressable.
For example, a 32 bit computer (that is, one which uses 32 bits for its memory-adresses) can typically adress up to 2^32 _bytes_ of memory (4GB), and not only, as you seem to think 2^32 _bits_ of memory.
The latter would make 512MB the maximal amount of RAM in a 32-bit x-86 machine.
In principle, a 32-bit machine could force aligned access, and thus only have one adress for every 4 bytes of memory, that would make it possible to adress 16GB with 32 bits. The reason this ain't done is probably mostly bacwards compatibility, I'm under the impression that most actual memory-access is 4-byte aligned anyway, for performance-reasons.
You're rigth. The armed public revolting against government in the USA won't happen, and could not suceed anyway, given the relative strengths of the military and any reasonable armed public. In 2003 the US military spent more money than the 12 next countries on the list combined. And it looks like in 2004 it'll be even higher.
It could happen, but only if either a) The US military itself or atleast a significant part thereof supported the revolution. Or b) at the very least the US military refused to shoot at the revolutionists.
The "armed population" is only relevant in b). If you can win the US-military over, effectively you're already running the country. If the US military would do what I say, and refuse to do what Bush say, I would be the effective president of the USA.
This is only partly true. *some*, perhaps even *most* cheats are a result of the fact that, as you say, with common game-designs the client knows a lot more than it *should* tell the player. And so by changing the client to disclose it, you gain an advantage.
Typical examples include cheats that let you see trough walls, or cheats that give you a wider field of view.
But that's not the *entire* problem. There's a few more classes of cheats. And those cannot be eliminated by the server only telling the client stuff that the player should see;
Aimbots. You see (and should see) the enemy on the screen. Some program helps you aim so that you hit better than you otherwise would. If you always hit 100%, it could conceivably be detected, but the problem is that the aimbot can be adjusted to be *precisely* as good as the server will allow anyone to be.
You mention casinos, card-games and turn-based games. Those can be cheated (well, it's up to you if you count it a cheat, but atleast it'll give one player an unfair advantage over the other players) for example by having a program count cards for you. It's quite a big advantage in for example online bridge to *always* know *exactly* which cards have been played and which remains. Good players will remember some of it, but a program will remember all 100%.
Then there's the problem of behind-the-scenes communications. To stick with the bridge-example, two players on a team have a *humongous* advantage if they can tell each others, somehow, what cards they have. With online gaming, this is obviously as simple as IM.
Or let's say online poker. Let's say it's implemented with Schneiers cryptographically secure poker-system, so that no client can cheat. But, the thing you don't know is that the three other guys at the table are really friends, and communicating over IM. They'll tell eachothers who has the best hand, and the others will fold. Essentially, you're playing against a player who gets three hands every round, and can choose the best one to play with. You will loose. There is nothing the game-client can do to prevent this. Even if it *somehow* blocked all other ip-communications, the others could be sitting in the same room and communicating by talking, or they could be sending eachothers sms or any of a 100 other possibilities.
You're rigth that telling the client less will reduce or eliminate *some* types of cheats. You are wrong however in claiming that this is the only reason cheating is possible at all.
You've been drinking the kool-aid. Fact is, in most jurisdiction it's not settled whether "eulas" and click-trough licenses are enforcable at all, in a few it's pretty clear they are not.
Imagine the following situation. I walk into a computer shop in, say, Norway, and say something along the lines of "I'd like a copy of Windows xp please.", the guy in the shop gives me a box and says something along the lines of "That'll be nok X then." I pay the price, take the product and leave.
What just occured was a *sale*. A fairly typical sale. It was not a "licensing", no reference where made to licenses of any type by me or the seller. A good assumption would be that, when I ask to purchase one copy of a copyrighted work, like XP is, that is, indeed what I get.
It is not reasonable to assume that after the above transaction, the customer should somehow magically know, and accept, that what he got was something completely different from what he asked for. MS would claim the customer had infact, not bougth a copy of XP, but instead paid for a license to use windows xp -- on certain terms.
But those terms where neither disclosed, nor mentioned at the sale. Infact, their very existence was not even hinted at. When they show up, at some later point, they are printed in english, a language my younger brother, for example, would not even understand while installing the thing. By which legal theory would I be bound by terms that I've never seen, and never heard about ?
Now, this does *not* mean I can do whatever I want with the work. It is copyrighted afterall, and copyright-law sets clear limits to what anyone other than the creator can do. I can, for example, not *copy* the work and resell the copies. Nor can I perform the work publically, and so on. (though norwegian law contains explicit permission to make one copy for backup-purposes, and to copy the program from the CD-rom and into the working-memory of the computer as nessecary to use the program.)
So, for many people, eulas truly are irrelevant. What *is* relevant when it comes to what you can and cannot do with a copyrigthed work is, you guessed it, copyright-law.
I also notice that you use tilted language. That's dumb and makes your argument look weaker than it is. Making extra copies of Windows is not "software theft", it is illegal copying. Breaking into a shop and stealing the windows-xp boxes stacked there would be software theft, and would be persecuted as theft. There is nothing common between theft-law and copyright-law, trying to mix them up only makes you look dull.
That's good. Your case could need it. I'll spoonfeed you again.
*ALL* I was saying was that the amount of water that goes trough a factory is not a useful measure of ecological impact. To be able to say something useful about the ecological impact, you'd have to say what degradation, if any, happens to this water.
Slashdot Mirror
That is actually very easy to answer.
When people commit crimes, we punish them. We do so for three reasons;
- In the hope that being punished for doing something wrong this time will prevent the person from doing it again. (individual preventive)
- In the hope that others, seeing that this person gets punished for doing something, avoids doing the same thing. (generally preventive)
- Protect society against the actions of a criminal by physically preventing him/her from doing it. (only in case of jail)
It is fairly obvious that atleast the first of these are pretty nonexistant for reach people with constant-sum fines. Saying that going 70mph where only 50 is allowed gets punished by $150 fine is the same as saying that anyone for whom that sum is without consequence can calmly go 70 all the time. Thus, the punishment is not working.More reasonable is a fixed percentage of your annual salary. The *reality* is that fining Joe Average $1000 is a significant deterrent for Joe, while to i.e. Gates it's completely irrelevant, it's not even interest on his fortune for the 5 minutes the cop hold him.)
Such a system does indeed exist atleast for some areas in Scandinavia. For example, in Finland if you speed, you get a fine based on your income. And for example in Norway, if you fail to file your tax-declarations timely, you're fined based on your income and wealth.
Assume the following:
- I download their drivers, direct from their homepage, and compile the module.
- Somewhat later I start to wonder what license the thing really is under.
- So I use modinfo as per the standard. When I do so, the license is listed as "GPL".
- This is the *intended* result. They *deliberately* made the string so that that would be displayed.
- I assume they mean what they say, and treat the module as if it is GPL.
I don't think they'd have a leg to stand on if they later wanted to try to punish me for breaking their license.Everything I did was perfectly reasonable. I was not in the least careless. The fact that I'm confused about their intended license is a *sole* result of the fact that they themselves *deliberately* misled me.
There's something called estoppel. It basically says a single entity cannot contradict itself to the detriment of others. Thus, one single company cannot at the same time deliberately have their modules print "GPL" when queried over the standard interface as to the license, and at the same time try to have me punished for believing that the module is in fact GPL.
For software-developers, the best thing to do about patents is to figth politically against them, and otherwise ignore them as best as possible. The *only* way to be certain you won't infringe some software-patent is never to release any software at all.
Sort of. But you have to look at the limits too, not only how "many" orders there are for buy and sell. If "Last" was 7, and there are tons of people who want to sell at 7, but hardly any buyers who want to pay more than say 5, it does indeed indicate the stock is likely to drop. If however, Last was 7, and there are *tons* of people who want to sell their stock at 9 while there are only a few people who want to buy at 8, this does not indicate a downward trend. I guess my point is, look at the price, not only the order-depth.
par2. It's a program mostly used for posting big binaries on usenet, but it works perfectly for this purpose. The basic idea is that you split the file into n+x pieces in such a way that any n pieces is sufficient to recreate the entire file. So, for example, assuming the recording of grandpa is 100MB, you use par2 to split it into (for example) 6 25MB pieces in such a way that *ANY* 4 of those pieces can be reassembled to the original file. Offcourse you could just make multiple complete copies, but that would take more space. the 6*25MB pieces take only 1.5 times as much place as the original 100MB, but you're still safe aslong as no more than 2 of those 6 pieces go lost or corrupted.
So, those 20 people who use bsd as a network *client* are somewhat less likely to have their tcp-connections successfully attacked as those who use predictable source-ports. (still not 48000 times safer as Theo writes, predictable does not typically equate "100% guessable with 1 try")
This "vulnerability" is kinda lame really. Previously, people who didn't think about it very much, assumed that since to reset a TCP-connection you need to guess the sequence-number, the chance of successfully doing so would be no higher than 1 in 2^32.
This "vulnerability" only points out that infact tcp-implementations will accept as valid any sequence-number between $CURRENT and $CURRENT+$WINDOW_SIZE.
So, instead of needing to try 2^32 times, you need "only" to try (2^32)/$WINDOW_SIZE times. Still fairly hard under typical conditions.
Window-size is however typically proportional to bandwith and inversely proportional to delay, so it'll be easiest to exploit on a tcp-connection that has high bandwith, and high ping-times. For example any connection that goes over satelite. (those of you that knee-jerk and think that high bandwith and high ping can't coexist should go reread first-year networking-curriculum.)
Your method works, unless your distro puts "DontZap" in as an option in the config. I think that for modern distros, the "approved" method is probably logging in as root and doing "service dm restart", but there's literally dozens of methods that will work, yours too, for most distros.
Afterall, all the install does is put some files at the apropriate locations, and change a few lines in a config-file.
X *does* however need to be restarted before it'll take *advantage* of the new drivers. Atleast I know of no way to tell a running X-server to go re-read it's config and re-initialise drivers.
No. But it'll less often accidentally delete stuff you really want than you would yourself if you manually waded thorough 1000 spams a day, attempting to find the dozen or so legitime email in between, and also trying to avoid wasting *Too* much time sorting the spam away.
For example, it's my opinion that reducing copyrigths to oh, say, 15 years, would massively help the progress of the useful arts. Yes, it'd screw over a few humongously rich people who're today sitting on the few copyrigthed items that are a) old, and b) worth something, but the purpose of copyrigth-law was never to benefit the MPAA, or anyone else. The purpose was, the only legal purpose by the constitution, to benefit the progress of science and the useful arts.
Noone is seriously suggesting that a significant number of players would cheat in this manner, only that the existence of this cheat-possibility demonstrates the (ultimate) futility of anti-cheat technology.
In the real world, a cheat can be software-only aslong as the user has complete control of his computer. The only way to prevent this is to take away the users control of the computer.
The principal provlem is that your game communicates with the world around it (including the graphics-card and the mouse and the network) by way of sending requests to the OS, and have them responded to. Aslong as the user controls the OS, he can have them responded to the way he likes, and there's nothing you can do about it.
In a computer the user controls, the OS can say: "mouse moved to x,y", and your game has no way whatsoever of verifying if this really happened, or if the OS is lying to you.
I don't know if you're willing to give up control of your computer and your OS and hand the keys over to "content creators", just in order to be able to enjoy a cheat-free online gaming experience. Personally I consider that a too high price to pay by something like 3 orders of magnitude.
It is different from a normal poker game because, not only are the three cooperating with eachother and sharing their pot (people could do this in normal poker), but they *also* has access to a perfect communication-channel where they can freely exchange information with no chance of it being discovered.
Whether you are able to see it or not, that *does* make a real difference. The principal difference is that they can always *know* for *sure* which of the three has the best hand, so who should fold, and who should bet.
The reason folding the two worse hands make a difference is because, though it doesn't influence their chanse of winning, or the payout when they do. It *decreases* their losses on the occasions when they loose.
Essentially, it is easier to guess correctly the answer to the question: "Do I have the best hand", when you get to see two of the other hands.
You're rigth about more chips being an advantage. That is why in poker-competitions you usually start of everyone with the same initial stack of chips.
- Speed in use. KDE seems (subjectively, I admit) to spend longer than XP in starting up, but to respond more snappily once it is up. To me that's a plus, because I don't constantly reboot my computer.
- Flexibility. For example, the new ability to easily write your own handlers for given mime-types is a goodsend. For example, you can now, easily, say that for files of type image/jpeg right-clicking the file should give the additional menu-entry "Show exif-information". Or that right-clicking a folder (inode/directory) shall give the option "backup to backupserver". Doing such things requires nothing more than that you are able to give a shell-script command that does the wanted stuff.
- Protocol-handlers. It's all nice and good that the file-manager can handle http:// aswell as file://, but with XP that's about where it ends. KDE 3.2 handles a lot more. cameras:// fish:// (access a host over ssh), smb:// and so on. Writing new kioslaves are a bit more tricky, but the collection is growing at a nice rate nevertheless.
- Automation. dcop makes it possible to control the entire gui from a script, and gives flexibility to configure kde so it responds the way *you* want it to. Want IM messages from someone received in kopete to automatically ask kweather to look up, and show you, weather-info for the place where your pal lives ? No problem.
- Freedom. Okay, so that's really mostly a feature of the license, and not a feature of the software, I happen to love it anyway. YMMV
That's a start. By the way, I'm not saying that XP doesn't do anything well. Maybe it does. You'll have to ask someone else about that though.If you ask for the data at "0000 0001" you'll instead get the stuff from "0000 0001" to "0000 0004", notice the overlap with the last call.
What this means is that there are 4 different ways of getting at every byte in memory. If you really want to read byte 17, you can do so by asking for adress 17, 16, 15 or 14, and discard those bytes you don't need. (this works because on 32-bit architectures, when you ask for "14", you really get 14-15-16-17, or atleast can at no extra delay. (because the bus doesn't go any faster with just 8 bit on it instead of 32 bits)
An alternative would be for a architecture to only be word adressable. This would mean that if you asked for adress 0 you get the first word in memory, which is bytes 0-1-2-3. If you ask for adress 1 you get bytes 4-5-6-7, in general, if you ask for address X you get word number X, which is bytes number X*4, X*4+1, X*4+2, X*4+3.
A "drawback" of this architecture would be that if you really wanted adress 17, you'd *have* to do it by asking for word 4, which would return you bytes 16-17-18-19, and then discard the unwanted stuff. Your CPU could have an opcode that did this for you though, and besides, we're pretty much doing this anyway, because frequently aligned (as in the adress I ask for is divisible by the number of bytes in a word) access is frequently faster than non-aligned.
You're rigth that sometimes there's "address bits" that aren't really part of the adress, so the effective adress-space is smaller. For example, for a long time Linux used 1 bit to determine if the address was "user space" or "kernel space", which effectively limited a single user-space program from having a adress-space larger than 2^32.
But with cheating even a complete newbie will know 100%. Besides, even the best bridge-couple in the country can't know 100%, from the get-go, what cards their partner has. Yes, we use the (wussname-in-english?) 'bidding' phase to communicate a bit to eachothers, but this is a very low bandwith channel, typical information is on the level of "I've got 13-20 points", if you've got strong cards and can bid high, you can transmit a bit more info, perhaps, for example, "Clover is my strongest colour", "I have few squares"
Knowing 100% of what your parner has (and thus 100% of what your two opponents have, only not who of them has what) is a humongous advantage.
Online play is still possible offcourse. But you'll be more or less forced to either accept constantly loosing to the cheating 12 year olds (which gets old real fast!), or do as I do -- play online only against people I know and trust.
The first computer would have no way of knowing that the commands came from a second computer, and not from a mouse held by a human.
*even* if you could somehow cryptographically ensure the integrity of the mouse, this would *still* gain you nothing, because there's nothing preventing a sufficiently determined person from creating a motorized trackball under the mouse that rotates under command of the second computer and thus creates mouse-movements.
The problem is analogous to the one of "copy-protecting" music, but still leaving the music hearable to the buyer. It's simply not possible, for the fairly fundamental reason that if you can hear it, so can a mic. Similarily, aimbot-proofing a online shooter is not possible because if a human could move a mouse like *so*, then the aimbot could do the same using mechanics.
You could say the mouse should use biometrics to ensure you're the one holding it. But a) I'm thinking you see that this is getting ridicolous and b) that also wouldn't help, I'd simply hold the mouse, and have a piece of the table move automatically under it so all I need to do is hold my hand more or less still.
Yes it will. You are being disingenious. I *said* that when I said "cheating" I really meant unfair advantage. Are you saying that getting three hands dealed, and getting to choose which one to play in poker is *not* an advantage ?
The fact that an advantage *can* theoretically sometimes be overcome by a much better player is irrelevant. That is like saying that doping is not cheating aslong as there exists someone who can beat you even when you are doped and they are not. That's nonsense.
Problem is, the problem *is* cheating. Not only that the people win a lot. People *do* care if the person who just grind them into a fine dust is really a super-player, or if it's the teenage-kid newbie from down the block with a +700% health-patch.
One of the points of training in a game, or sport, is to be able to see yourself getting better, to increase your reputation.
Who would bother spending a lot of time becoming the best possible CS-player if all that means is that you'll be able to compete successfully in a group consisting of the newbie from last week that upped his health to ten times the normal ? Would that still be a challenge and fun ? Even when you know that what you achieve with an extra half-year of training, the kid will compensate by multiplying his health not by 10, but by 25 ?
Sorry, I just don't believe in your thesis. That people only care about an even match, and *don't* care at all about fairness and cheating. Handicaps is something different, the principal difference being the openness. There's a difference between saying "Mr Wood, I challenge you to a game, I recognize you are much better than me, so I propose we add 30 strokes to your result, and see if I can beat that." and saying "Mr Wood, I can kick your butt any day of the week" (because I have a cheat installed, but I won't say that.)
What they *do* point at, pretty consistently, is a less than reputable, less than serious company. Not that it's a huge surprise that SCO has dealings with consultants of similar moral fiber.
As long as a vendor knows that you will buy his shit no matter what he demands for it, no matter how late it is, and no matter how it performs, he has no incentive to charge less or deliver more.
As a customer, you have zero negotiation-room aslong as both you and the vendor knows that he's got you by the balls.
His ball-grip *only* looses at the point where you seriously start to consider other vendors. What would your company if the next version of Windows cost $500/processor ? Suck it up or change vendors ? $1000 ? $10.000 ? Point is, sooner or later there comes a point where you'll say "No. I won't pay that. Sorry."
Going to a convicted monopolist and saying the equivalent of "I'll buy your stuff at ANY price, but could you please sell it to me cheaper anyway ?" is a nonstarter.
These days, however, going to Micrsoft and saying "Look, upgrading to the newest Office for our 700 desktops is going to cost $X, we like Office, mostly, but the thing is, we also got an offer for something called OpenOffice, and it seems that even with retraining-costs it'll still only be $0.5X, and besides, the upgrades thereafter are free." has a tendency to get you offered 50-90% discounts on the MS-software.
This demonstrates perfectly *exactly* where the ball-grip has it's roots. It exists only when Microsoft feels sure you'll buy their stuff anyway.
For example, a 32 bit computer (that is, one which uses 32 bits for its memory-adresses) can typically adress up to 2^32 _bytes_ of memory (4GB), and not only, as you seem to think 2^32 _bits_ of memory.
The latter would make 512MB the maximal amount of RAM in a 32-bit x-86 machine.
In principle, a 32-bit machine could force aligned access, and thus only have one adress for every 4 bytes of memory, that would make it possible to adress 16GB with 32 bits. The reason this ain't done is probably mostly bacwards compatibility, I'm under the impression that most actual memory-access is 4-byte aligned anyway, for performance-reasons.
This ain't anywhere close. It's not even using a hundreth of a percent of the 64 bit adress-space
It could happen, but only if either a) The US military itself or atleast a significant part thereof supported the revolution. Or b) at the very least the US military refused to shoot at the revolutionists.
The "armed population" is only relevant in b). If you can win the US-military over, effectively you're already running the country. If the US military would do what I say, and refuse to do what Bush say, I would be the effective president of the USA.
Typical examples include cheats that let you see trough walls, or cheats that give you a wider field of view.
But that's not the *entire* problem. There's a few more classes of cheats. And those cannot be eliminated by the server only telling the client stuff that the player should see;
Aimbots. You see (and should see) the enemy on the screen. Some program helps you aim so that you hit better than you otherwise would. If you always hit 100%, it could conceivably be detected, but the problem is that the aimbot can be adjusted to be *precisely* as good as the server will allow anyone to be.
You mention casinos, card-games and turn-based games. Those can be cheated (well, it's up to you if you count it a cheat, but atleast it'll give one player an unfair advantage over the other players) for example by having a program count cards for you. It's quite a big advantage in for example online bridge to *always* know *exactly* which cards have been played and which remains. Good players will remember some of it, but a program will remember all 100%.
Then there's the problem of behind-the-scenes communications. To stick with the bridge-example, two players on a team have a *humongous* advantage if they can tell each others, somehow, what cards they have. With online gaming, this is obviously as simple as IM.
Or let's say online poker. Let's say it's implemented with Schneiers cryptographically secure poker-system, so that no client can cheat. But, the thing you don't know is that the three other guys at the table are really friends, and communicating over IM. They'll tell eachothers who has the best hand, and the others will fold. Essentially, you're playing against a player who gets three hands every round, and can choose the best one to play with. You will loose. There is nothing the game-client can do to prevent this. Even if it *somehow* blocked all other ip-communications, the others could be sitting in the same room and communicating by talking, or they could be sending eachothers sms or any of a 100 other possibilities.
You're rigth that telling the client less will reduce or eliminate *some* types of cheats. You are wrong however in claiming that this is the only reason cheating is possible at all.
Imagine the following situation. I walk into a computer shop in, say, Norway, and say something along the lines of "I'd like a copy of Windows xp please.", the guy in the shop gives me a box and says something along the lines of "That'll be nok X then." I pay the price, take the product and leave.
What just occured was a *sale*. A fairly typical sale. It was not a "licensing", no reference where made to licenses of any type by me or the seller. A good assumption would be that, when I ask to purchase one copy of a copyrighted work, like XP is, that is, indeed what I get.
It is not reasonable to assume that after the above transaction, the customer should somehow magically know, and accept, that what he got was something completely different from what he asked for. MS would claim the customer had infact, not bougth a copy of XP, but instead paid for a license to use windows xp -- on certain terms.
But those terms where neither disclosed, nor mentioned at the sale. Infact, their very existence was not even hinted at. When they show up, at some later point, they are printed in english, a language my younger brother, for example, would not even understand while installing the thing. By which legal theory would I be bound by terms that I've never seen, and never heard about ?
Now, this does *not* mean I can do whatever I want with the work. It is copyrighted afterall, and copyright-law sets clear limits to what anyone other than the creator can do. I can, for example, not *copy* the work and resell the copies. Nor can I perform the work publically, and so on. (though norwegian law contains explicit permission to make one copy for backup-purposes, and to copy the program from the CD-rom and into the working-memory of the computer as nessecary to use the program.)
So, for many people, eulas truly are irrelevant. What *is* relevant when it comes to what you can and cannot do with a copyrigthed work is, you guessed it, copyright-law.
I also notice that you use tilted language. That's dumb and makes your argument look weaker than it is. Making extra copies of Windows is not "software theft", it is illegal copying. Breaking into a shop and stealing the windows-xp boxes stacked there would be software theft, and would be persecuted as theft. There is nothing common between theft-law and copyright-law, trying to mix them up only makes you look dull.
*ALL* I was saying was that the amount of water that goes trough a factory is not a useful measure of ecological impact. To be able to say something useful about the ecological impact, you'd have to say what degradation, if any, happens to this water.