Picture this from the point of view of the officer.
You stop someone who looks like a trouble maker. Perhaps you suspect him of drug dealing, but probably not. You ask him what's in his vest, which IS WITHIN YOUR RIGHT AS A GENERAL CITIZEN. He just doesn't have to answer. Instead he says:
"Left my 'nades at home, officer."
Now, by his own admission, he had
pens, granola bars, cigarettes and notebooks
in his vest. Through the vest, these look exactly like granades. And, since poor negative use and sarcasam are common in English, his statment could easily be taken as "I'm a wacko with a vest of illegal weapons!" And from the cop's point of view, it looks like he is (remember how he's dressed).
If I were the cop, he would be getting aquainted with the pavement and Mr. Colt untill I could have him strip searched. The cop acted with INCREDIBLE restraint in the face of potential personal injury or death. And this guy has the gall to complain about the cops violating HIS rights...
... we need to admit that they're right. The thing is something like a year behind schedule, and look what it delivers: INCREDIBLE lack of speed pretty large bloat good standards no JAVA on our fave OS lots of bugs now, compare to konqueror - starts LITERALLY 50 times faster on py p2-266 less bloat same standards any jdk you want slightly more bugs Now, should we as open source developers be trying to fix the major problems above, or the bugs below. Get on the RIGHT bandwagon.
Ok this is not a trivial problem, because Arizona managed to screw it up to a certain degree, but it can be done. The MIT grad (I hope he graduated) above hinted that good protocols existed, and here's an example off the top of my head:
Step 1: A long random bit field is given to each voter via mail. This is no more or less secure than vote by mail.
Step 2: The vote counter gets this field also.
Step 3: Voter goes to the poll page, and submits the first half of the key to the poll system 1. It is matched against the first halves of all the keys in the system. No match -> no vote else match -> then send the second half of the key to a second vote system.
Step 4: Delete the entire key on the first system.
Step 5: Voter uses second half of key to authenticate with poll computer 2.
Step 6: Voter submits vote (protected by ssl or whatever) to system 2. Vote is paired with second half of key, and tallied.
Step 7: Second half of key is posted in public.
Step 8: If truely paranoid, destroy computer 1.
Result: Voter is anonymous. Voter is authenticated at levels better than what is curently in use. Dead people cannot vote unless they can get mail, which is more expensive than the current vote-the-dead schemes. Voter can prove the voted by matching second half of key to list of published keys that voted. No one else can prove that you voted. You can't prove how you voted, so you can't be bribed.
Executive Summary: Faster and more secure than current system.
Grab a copy of the second edition of shadowrun. It has a "hitory of the future" section just like the third edition, but it was published far earlier ('93?). Then compare it to the 3rd edition. What you'll find is that a large number of events were taken out becaus ethey already happened.
Some were obvious, such as the colapse of the old Sout African system, but the sheer number they got right is astounding.
Examples: First brain-computer interface First cybernetic vision system First implanted arm/leg
The reality it that witht he exception of the magic crap, shadowrun predicts the future amayzingly well.
In a stunning decision, judge Grand Pohbah handed down a ruling punishing M$ for their illegal tying of words. Analyst: M$'s strategy is simple: first they embrace a word - say "Innovate." Then they start to extend it, adding "right to" and making it apply to things that previously never fell under the header of "inovate." The resulting "Inovation" is incompatible with anyone else's, but by that point the M$ inovation is a defacto standard, and OSM (Origional Sentance Manufacturers) are tied into the scheme.
IANAL, but I've advised out leagal team in the breaking of 3 tech patents via prior art, so I know this stuff.
All the "Evil Patent" articles ignore the point. The USPTO grants all kind of stupid patents because it's not fundamentally their job to determine if the patent is valid. All they do is run down a procedural checklist and grant the thing. If they find a glaring example of prior art, they might reject it, but usually not.
Where prior art comes in to play is in the enforcement part.
Example: CNET makes the patent. I ignore it and violate it right and left. CNET sues to stop me. Now, my defense will be my examples of prior art. And in the case of almos all these tech patents, it would be a very sucessfull defense.
The moral: This patent is so weak I (having watched the trials, but without training) could break it with 90%+ confidence. A real lawyer could do it 98%+
On FreeBSD, the Linux emulation is good enough that I run Blackdown 1.2.2rc4 without trouble. Just install it from the port, not by hand, since there's no reason to re-invent the wheel. With the better JIT, it's faster than the native port anyways (although the freebsd-java project is changing that). In fact, the only thing I've ever had fail under emulation was a port of linux wine, and that just needed linux-proc, which wasn't installed. The whole emulation thing is rock-solid.
Unfortunatly, I'm under NDA, so I can't tell you everything about it, but I have a fairly good grasp on the performance of this class of device.
The biggest problem with 32 bit machines is not the 32 bit int, which is really sufficient for most things, it's the inability to address more than 4 gig of memory. This provides a relatively clean solution to that problem by using this device as swap. The burning question, of course, is performance... how much worse is it than on board memory?
The answer is, of corse, it depends: If you are in a single process enviroment, the time it takes to swap pages is somewhat killer, because the machine justs sits on it's ass while the DMA moves the block. Now, don't get me wrong, it's a lot better than disk, but it's not like real meory.
However, on multi-process machines like servers, it's great. There is a delay for the page swap, but the other processes keep the cpu busy and the DMA keeps the bus busy. Since throughput is more important than response time, this is almost as good as onboard ram. But, you say, this is MORE expensive than real RAM? not really... for an app like this, it will be an smp machine anyways, and the difference in cost between comodity x86 parts and a 64bit+8gig-uberboard setup from a proprietary vendor is so great that you could buy one of these things with the spare change. This could easily save many tens of thousands on certain types of server projects.
Some manager I had said: If you want to be sucessfull, find a sucess and see how it was made.
The obvious canidate would be Bill Joy's TCP/IP implementation. Eveyone runs it:
1. BSD's always used it
2. SYS V incorperated it - thus it flowed to most commercial unixes
3. LINUX borrowed heavily from is (recall that Regents of the University of California boot message?)
4. If the TCP/IP fingerprint of WIN2000 is any indication, they borowed it too.
And it works right every sincle time you use it. So, what process made it? A single genius. All the cool process in the world won't make up for the fact that the single requirement for great software is a great designer/programmer. The required process is simple - whatever that person requires to let their genius loose.
The only way to circumvent this requirement is to do what NASA does and spend probably literally hundreds of $ per line of code.
"Group we loves looses free-speech/privacy/fair-use case. This sucks..."
Of course, what this leaves out is that these types of cases eventually make their way to the supreme court, and that all the prior decisions are 100% unimportant. Only that last trial counts. Everything else is just there to convince the court that the case is important enough to take. Now don't take me wrong, the Napster case itself may not make the grade, but eventually a test case will come before the court.
Now, the court hates to reverse itself. That means that their decision will be based on other similar cases, such as the VCR. And we all know that the decision on VCRs was that if it had a single legal use, it was legal. And since the court almost certaily won't reverse itself, Napster is safe as can be. Don't let the RIAA news blurbs fool you!
Why? Because we're told not to!
on
Why Not Ada?
·
· Score: 1
I'm a CS student at CU Boulder. I'm lucky enough to get to choose my language for almost any of the projects I work on. Of course, the instructors try to give us hints on what to use. Whithout anyone asking about Ada, almost every instructor chooses to complain about it at some point. What am I as a student to do? Should I learn it in favor of tons of other languages that are recomended, popular, used in the real world, and that employers want me to know?
I once tried to get a Linux box past the boss through "legit" channels, and had a major success. We were replacing an older-than-god Sun mail server, and I suggested a Linux box. At the time I think it was Slackware. Got it all set up, we moved it inot the network, and it worked fine. However, the boss decided to cover his ass, and bought an NT server and a commercial mail program that will remain nameless (you'll see why in a bit). I was miffed, but rather than sulk or smuggle, I got out the hex editor and disassembler. Two hours later, I had found 10 unbounded str* functions that lead to buffer overflows. Wrote up an exploit, and showed it to the boss. He didn't really believe it, but let me run the thing, and sure enough, it worked. Two hours and a little help from me and the now-classic AlephOne article later, he had written his own exploit on a different hole. At that point he sent the mail program back, demanded a refund, and there's a linux server there to this day.
Slashdot Mirror
You stop someone who looks like a trouble maker. Perhaps you suspect him of drug dealing, but probably not. You ask him what's in his vest, which IS WITHIN YOUR RIGHT AS A GENERAL CITIZEN. He just doesn't have to answer. Instead he says:
"Left my 'nades at home, officer."
Now, by his own admission, he had
pens, granola bars, cigarettes and notebooks
in his vest. Through the vest, these look exactly like granades. And, since poor negative use and sarcasam are common in English, his statment could easily be taken as "I'm a wacko with a vest of illegal weapons!" And from the cop's point of view, it looks like he is (remember how he's dressed).
If I were the cop, he would be getting aquainted with the pavement and Mr. Colt untill I could have him strip searched. The cop acted with INCREDIBLE restraint in the face of potential personal injury or death. And this guy has the gall to complain about the cops violating HIS rights...
... we need to admit that they're right. The thing is something like a year behind schedule, and look what it delivers: INCREDIBLE lack of speed pretty large bloat good standards no JAVA on our fave OS lots of bugs now, compare to konqueror - starts LITERALLY 50 times faster on py p2-266 less bloat same standards any jdk you want slightly more bugs Now, should we as open source developers be trying to fix the major problems above, or the bugs below. Get on the RIGHT bandwagon.
Ok this is not a trivial problem, because Arizona managed to screw it up to a certain degree, but it can be done. The MIT grad (I hope he graduated) above hinted that good protocols existed, and here's an example off the top of my head:
Step 1: A long random bit field is given to each voter via mail. This is no more or less secure than vote by mail.
Step 2: The vote counter gets this field also.
Step 3: Voter goes to the poll page, and submits the first half of the key to the poll system 1. It is matched against the first halves of all the keys in the system. No match -> no vote else
match -> then send the second half of the key to a second vote system.
Step 4: Delete the entire key on the first system.
Step 5: Voter uses second half of key to authenticate with poll computer 2.
Step 6: Voter submits vote (protected by ssl or whatever) to system 2. Vote is paired with second half of key, and tallied.
Step 7: Second half of key is posted in public.
Step 8: If truely paranoid, destroy computer 1.
Result: Voter is anonymous. Voter is authenticated at levels better than what is curently in use. Dead people cannot vote unless they can get mail, which is more expensive than the current vote-the-dead schemes. Voter can prove the voted by matching second half of key to list of published keys that voted. No one else can prove that you voted. You can't prove how you voted, so you can't be bribed.
Executive Summary: Faster and more secure than current system.
Some were obvious, such as the colapse of the old Sout African system, but the sheer number they got right is astounding.
Examples:
First brain-computer interface
First cybernetic vision system
First implanted arm/leg
The reality it that witht he exception of the magic crap, shadowrun predicts the future amayzingly well.
In a stunning decision, judge Grand Pohbah handed down a ruling punishing M$ for their illegal tying of words. Analyst: M$'s strategy is simple: first they embrace a word - say "Innovate." Then they start to extend it, adding "right to" and making it apply to things that previously never fell under the header of "inovate." The resulting "Inovation" is incompatible with anyone else's, but by that point the M$ inovation is a defacto standard, and OSM (Origional Sentance Manufacturers) are tied into the scheme.
All the "Evil Patent" articles ignore the point. The USPTO grants all kind of stupid patents because it's not fundamentally their job to determine if the patent is valid. All they do is run down a procedural checklist and grant the thing. If they find a glaring example of prior art, they might reject it, but usually not.
Where prior art comes in to play is in the enforcement part.
Example:
CNET makes the patent.
I ignore it and violate it right and left.
CNET sues to stop me.
Now, my defense will be my examples of prior art. And in the case of almos all these tech patents, it would be a very sucessfull defense.
The moral: This patent is so weak I (having watched the trials, but without training) could break it with 90%+ confidence. A real lawyer could do it 98%+
On FreeBSD, the Linux emulation is good enough that I run Blackdown 1.2.2rc4 without trouble. Just install it from the port, not by hand, since there's no reason to re-invent the wheel. With the better JIT, it's faster than the native port anyways (although the freebsd-java project is changing that). In fact, the only thing I've ever had fail under emulation was a port of linux wine, and that just needed linux-proc, which wasn't installed. The whole emulation thing is rock-solid.
The biggest problem with 32 bit machines is not the 32 bit int, which is really sufficient for most things, it's the inability to address more than 4 gig of memory. This provides a relatively clean solution to that problem by using this device as swap. The burning question, of course, is performance... how much worse is it than on board memory?
The answer is, of corse, it depends: If you are in a single process enviroment, the time it takes to swap pages is somewhat killer, because the machine justs sits on it's ass while the DMA moves the block. Now, don't get me wrong, it's a lot better than disk, but it's not like real meory.
However, on multi-process machines like servers, it's great. There is a delay for the page swap, but the other processes keep the cpu busy and the DMA keeps the bus busy. Since throughput is more important than response time, this is almost as good as onboard ram. But, you say, this is MORE expensive than real RAM? not really... for an app like this, it will be an smp machine anyways, and the difference in cost between comodity x86 parts and a 64bit+8gig-uberboard setup from a proprietary vendor is so great that you could buy one of these things with the spare change. This could easily save many tens of thousands on certain types of server projects.
The obvious canidate would be Bill Joy's TCP/IP implementation. Eveyone runs it:
1. BSD's always used it
2. SYS V incorperated it - thus it flowed to most commercial unixes
3. LINUX borrowed heavily from is (recall that Regents of the University of California boot message?)
4. If the TCP/IP fingerprint of WIN2000 is any indication, they borowed it too.
And it works right every sincle time you use it. So, what process made it? A single genius. All the cool process in the world won't make up for the fact that the single requirement for great software is a great designer/programmer. The required process is simple - whatever that person requires to let their genius loose.
The only way to circumvent this requirement is to do what NASA does and spend probably literally hundreds of $ per line of code.
"Group we loves looses free-speech/privacy/fair-use case. This sucks..."
Of course, what this leaves out is that these types of cases eventually make their way to the supreme court, and that all the prior decisions are 100% unimportant. Only that last trial counts. Everything else is just there to convince the court that the case is important enough to take. Now don't take me wrong, the Napster case itself may not make the grade, but eventually a test case will come before the court.
Now, the court hates to reverse itself. That means that their decision will be based on other similar cases, such as the VCR. And we all know that the decision on VCRs was that if it had a single legal use, it was legal. And since the court almost certaily won't reverse itself, Napster is safe as can be. Don't let the RIAA news blurbs fool you!
I'm a CS student at CU Boulder. I'm lucky enough to get to choose my language for almost any of the projects I work on. Of course, the instructors try to give us hints on what to use. Whithout anyone asking about Ada, almost every instructor chooses to complain about it at some point. What am I as a student to do? Should I learn it in favor of tons of other languages that are recomended, popular, used in the real world, and that employers want me to know?
I once tried to get a Linux box past the boss through "legit" channels, and had a major success. We were replacing an older-than-god Sun mail server, and I suggested a Linux box. At the time I think it was Slackware. Got it all set up, we moved it inot the network, and it worked fine. However, the boss decided to cover his ass, and bought an NT server and a commercial mail program that will remain nameless (you'll see why in a bit). I was miffed, but rather than sulk or smuggle, I got out the hex editor and disassembler. Two hours later, I had found 10 unbounded str* functions that lead to buffer overflows. Wrote up an exploit, and showed it to the boss. He didn't really believe it, but let me run the thing, and sure enough, it worked. Two hours and a little help from me and the now-classic AlephOne article later, he had written his own exploit on a different hole. At that point he sent the mail program back, demanded a refund, and there's a linux server there to this day.