What exactly are the other implications besides copyright that you're talking about? What new laws will help reward artists?
Hmm, let's see:
More at stake: the power to use technology you purchase in an way you see fit (allowing people to directly maximize their experience of several important individual liberties, not least of which is the freedom of speech).
What new laws will help reward artists: Gee, how about a tax credit for payment to artists and scientists?
Those lawmakers would probably say that the gun/general-purpose-computer-plus-Internet analogy is incorrect and that the correct analogy is something along the lines of:
gun: nuclear device:: VCR: general purpose computer + broadband
I.e., while they advocate ownership rights of average citizens to a device whose beneficial uses are as broad as hunting, securing one's home, and contributing to the common defense -- and whose harmful uses are at worst the murder of a handful of people (before the police or some off-duty postman blow you away), they would advocate strict government control of a device whose benefits are undisputed, but who harmful effects can change the face of the world (their logic on all counts, not mine).
They teach auto shop in high school. Of course, fixing cars is not as "essential" as reading, writing and math, either. However, not *all* students will go on to college. Most will probably own a car at some point. For some, getting a job fixing cars out of high school may be an attractive option. So knowing how to do simple repairs on a car is a very *practical* thing for *many* high school students to learn, and I think that a lot of people would argue that therefore it's a valuable addition to the curriculum.
Many kids will either own computers or work with them daily after high school. Some may want to go on to work in an explicitly computer-oriented career, which however doesn't require much specific post-secondary education (hardware repair in a small shop, for instance). To the same degree as auto-mechanics (and probably far more than say, wood shop), computer education in secondary schools is a valuable addition to the curriculum.
For primary education, heavily computer-centric instruction may be overkill. But at the high school level good arguments can be made for it.
Of course, it won't be too useful to students who just want their school to subsidize their bong-building activities, but that's what metal shop is for.
Besides which, in any school you can find technologically savvy kids. Make them a part of the computer team that maintains the network.
In this age of lawsuit-happy parents and grandstanding local politicians, I don't see many school board members or superintendants who'd likely wan't to be associated with the idea of kids controlling their own use of technology, however flawed the idea of controlling access to technology in schools may be.
Few of the companies I've worked for, as a contractor *or* an employee, have had mandatory code review as part of the "official" process. Of the one or two that did, neither had any mechanism to determine whether the code was actually reviewed, and at both of them, especially as deadlines got nearer, code reviews were frequently not done even though everyone agreed they were supposed to be.
Is this good? NO! Is it common? In my experience, in the literature of our industry, and of the opinion of most of the programmers I personally know (which is a large sample), it is the *rule* rather than the exception, unfotunately.
it's embarrassing that this pseudo-technical forum is giving the report even a little credit
Well not all of us have the "benefit" of having worked at Microsoft, so you'll have to forgive us if we're ignorant as to the exact procedures used to test code there. Additionally:
For the months before the OS ships every line of code that is modified is examined on several levels; every bug that is found could potentially be investigated by any of dozens of people in any part of the organization...
OK. Do you have any description of this process that's more concrete than "examined on several levels?" Are all of the "any of dozens of people in any part of the organization" programmers? Is their level of debugging skill equal to or greater than the skill of the most skilled programmer producing the code? How many people investigate the bugs that aren't found? (Careful, this last one is a trick question)
There's nearly a 1/1 ratio of Test/Dev in the critical parts of the system; to do this you would have to get the developer(s) and the tester(s) responsible for that chunk of code/functionality.
OK, so assume *one* chunk is targeted. Does that mean they'd have to get *one* programmer/tester pair? Two? Is it inconceivable that there is a piece of the system advanced enough that there are only a couple people skilled enough to work on it? Is it conceivable that, in such a case, a *large* enough amount of money exists to corrupt those small # of individuals? What criteria are used to determine which pieces of the system are "critical?" Is browser code which allows files with spoofed content/file types to execute on the machine an example of this type of "critical code?"
Automated tools run by seperate groups review changes and record owners; try to sabotage something once & you won't get a second chance.
And these automated tools can distinguish between bugs and deliberate vulnerabilities, perfectly? Is the person writing the tools more skilled than the most skilled programmer writing the production code? What tests are done on the testing tools to determine that they work adequately?
I hope you get my drift. Do I personally believe that terrorists have infiltrated Microsoft and planted bugs in the code? Not likely. Is the scenario conceivable? Absolutely.
AFAIK, GA doesn't necessarily mean "self-modifying-at-runtime" code. It can merely mean code included in the finished product was selected due to its "fitness" to some purpose, using a process where successive iterations were performed and the best candidates chosen, altered using some form of tranformation akin to "mutation" and "recombination" and used as input to the next iteration.
I.e., while the code was being modified during the GA process, the result is the most "fit" bit across all iterations, included statically in the final program.
For example:
"I wonder what set of parameters can best be used to tune this algorithm -- A1 -- to produce more accurate results more quickly"
assemble a list of sets of parameters, and some algorithm -- A2 -- which runs A1 parameterized with each set of parameters in the list, selects the 5 that worked best, records that, creates a new list of sets constructed from bits of the previous winners
iterate n times
use the winning set of parameters over all iterations as the parameters to A1 in codebase C1.
the chance that they would be able to intentionally slip in a trojan horse without it being caught in testing are pretty low.
What software company did you work at where there was a test protocol in place to test for trojan horses? What was the protocol? Did they feed the program every possible combination of random input to see if it produced strange behavior (rather than just testing one or two instances of invalid input to see if an error was generated and/or handled gracefully)? Did someone have a list of all possible code-words a terrorist might use as input to a routine which did something seemingly innocuous like opening a temporary file for reading, while running with heightened permissions?
In a million-plus line codebase for a product under deadline pressure, while official policy might be that "every line is checked", in reality this is highly unlikely to happen. The coders and their managers may assure the suits, "Yeah, we reviewd every line of code," but they'd be lying. It just doesn't happen. It's one of those things that everyone knows is *supposed* to happen and most people know doesn't *really* happen.
Secondly, while I agree that it's unlikely that a terrorist would approach a 13-year old kid and say, "Hey, you should start excelling in Math and then attend college to get a CS degree so that 10 years from now you can go work at Microsoft for 4 years or so (enough to gain the confidence of your managers) and then start putting back doors and bugs in their OS," it's far more plausible that a terrorist would approach a already working programmer who's naive and idealistic -- and perhaps *already* working at and trusted by managers at Microsoft -- and say, "Hey, here's how you can really help your faith..."
That is why the meaning of words or phrases change over time. "Steep Learning Curve" means one thing technically in psychology, but another altogether in popular usage. Which version is correct depends on one's audience, not one's bias. If anything, it's likely that the meaning in popular usage will come to so dominate the original technical meaning, that psycologists will actually reverse the axes of the graph in the traditional examples and the old technical usage will become archaic.
No, it's not true. You, as the copyright holder, can distribute the product under any terms you see fit. People to whom you distributed it under the terms of the GPL, get to use it under those terms. People to whom you've distrubuted it under alternative license terms get to use it under those terms.
The only issues that arise are when other people start to have some copyright interest in your product; then, in order to distribute it under proprietary terms, you must have their permission as well.
This is why, for instance, Sleepycat Software requires patch submitters to assign the copyright for the patch to Sleepycat before they apply the patch to the official Berkeley DB code. Something like that approach seems neccessary unless you want to end up maintaining two codebases: the GPL one and the proprietary one.
"Simply download and look for exploits" is a double-edged sword, my friend; yes, crackers can do it, but so can security consultants & researchers. If the source is widely enough distributed, then the chances of both "black hats", "white hats" and the J-Radom-I-need-this-new-feature hacker discovering a bug are increased. Fortunately, if the "white hats" or J-Random-tinkerer find it, it can be fixed so that it's (mostly) irrelevant that the black hats also found it.
In the second, binary only scenario, the PITA of reverse engineering ensures that a much smaller # of both white and black hats will attach themselves to the problem, and J-Random-tinkerer can't contribute at all. You've made it less likely that some "bad" person will use the bug to hurt you, but you've also made it less likely that some "good" person will find it and help you avoid or minimize the damage from the "bad" person's discovery.
My point is, security is a gamble. You are always gambling that there are no bugs in your program that some bad person is going to discover and exploit to harm you. "Security through Obsurity" is the wishful thinking strategy; it seeks to minimize the # of people who find the bugs, and then hope that those people are only the "good guys." Open Source and full disclosure are the "hedge your bets" strategies. They seek to maximize the number of people who find the bug, hoping that at least *one* of them is one of the good guys.
Purely statistically speaking, which do you think is likely to be more successful?
88 + 167 = 255; In other words, the CNN Money article seems to be using the same figures you quoted previously (probably from a common source which is geared more toward PR than rigorous accounting).
Are promotional costs for overseas distribution covered under the "Budget" figure in that graph? I have heard it repeated that Waterworld "probably" broke even when overseas distribution was taken into account (even the link you posted says 'probably'). However, the 88 + 167 > 175 (which are the correct figures, BTW) formula is, I believe, naive.
It's probably more like: 88 + ((167 * x) - y) - 175 = n
where
x = percentage of gross recouped by studios (after all, movie chains bid on pictures, and AFAIK, that's what the studio gets. So in the case of a US dog, where the foreign theaters say "based on poor performance in the US, I'll only pay you $50 million", and then the film grosses $167 million, it's not clear to me who gets the bulk of the money); and
y = the cost of promoting the film overseas
I'd say that only the studio accountants really know whether n > 0.
They did tests on samples which it turned out were from seed virus lot, not production vaccine. The tests were worthless from the point of view of establishing whether or not actual vaccine production lots were contaminated with simian HIV precursor virus.
Additionally, there is now evidence that Chimpanzee tissue culture -- Chimpanzee SIV being the supposed AIDS precursor virus -- was used to passage the seed virus to produce vaccine (something which every opponent of the OPV/AIDS hypothesis claims never happened, but of whom none can prove, given that no protocol for vaccine production has ever been produced for the Koprowski/Wistar vaccines in question), and circumstantial evidence that vaccine produced in this manner may actually have been used in field trials in humans.
So,
The people who are claiming that the debate is over are the people with a compelling financial interest in squashing this line of investigation;
The people who are claiming that the debate is over are ignoring evidence as they see fit, and are content not to ask for evidence (e.g., vaccine manufacture protocols and lab notes related actual vaccine production lots) which would shed a great deal of light on the issue;
The proponents of the theory have no financial interest in the outcome of the debate(even if one could argue book sales, the books will sell even if the theory turns out to be false);
ONLY THE OPV/AIDS theory accounts for the extremely high correlation of mass polio vaccinations and early AIDS infection sites, especially in remote, supposedly "isolated" areas in Eastern Congo, Western Rwanda/Burundi;
While I am pro-science, I am strongly pro-caution. We all want to rid our species of diseases, but we must carefully weigh the potential costs with the uncertain benefits and strive to reach the utmost level of certainty with regard to safety. Unfortunately, the rush to be the first to do something (and thus get the profits associated with commercial development, and secure future funding) can be highly detrimental to that goal.
Apart from that, you hit the nail on the head -- non-netizens have spent their entire lives learning not to think about the information with which they're presented.
Netizens, exposed daily to the most fascinating truths, the wildest conspiracy theories, and the most mind-bogglingly-high piles of utter bullshit, have, by virtue of our existence, been forced to learn how to think critically about every piece of information we're offered.
You bring up an interesting twist.
On the one hand, I think that some people are just inherently critical thinkers -- I lump myself and many of my friends into that category. Whether this is innate, or due to experiences growing up (i.e., learning to distrust authority) is anyone's guess.
However, I think you may be right that context has something to do with it. We who spend lots of time on the internet have learned, as you point out, that the vast majority of information there is crap, of little value. Likewise, Joe Sixpack may be more critical of Gary Condit (because it's conventional wisdom that politicians lie, especially about sex) than "that feller my sister's friend's friend knows that had his kidney stolen during Mardis Gras."
Have to disagree with you on this. Whether the information is true or not is not the issue. The more you take in, the more informed you are.
That assumes that the "you" you're referring to is capable of critical thought. In my experience, the great majority of people, in the U.S., at least, don't have very well-developed critical skills.
Mind you, I don't believe that this is because people are "stupid." Our educational system, with its emphasis on rote memorization and deference to Authority, does a poor job of teaching people to evaluate information that claims to be "authoritative." For example, how many of those reading this post have received the "GOOD LORD, CONGRESS IS GOING TO TAX EMAIL!" chain spam from countless of their relatives and friends -- a hoax email which, due to the fact that it claims a non-existent Senator as the sponsor of a bill which breaks the pattern of identifiers for senate bills (i.e., doesn't start with SB) -- is easily identifiable as such?
Face it; most people lack the ability to evaluate the value of information on the Internet.
Therefore, I should get a patent. Yeah the frombozing people might get upset and say the fractilizing is a invfringement but that would be a tough call.. there are many inventions that work in a similar fashion, that may create similar items but are different because... well the designers made them different enough so they could get a patent.
Right. That's the crux of the issue. It's like Interface Inheritance vs Implementation Inheritance in OO programming languages.
If your invention inherits the implementation of the other invention, you need to license the other invention. If your invention merely inherits the interface (i.e., it produces similar results to the original invention, but without using the components -- be they physical parts, or the steps in a patented process), then you don't have to license the original. In the latter case, it's unclear to me though whether this is a "patented improvement" per se, or just a new patent.
As you said, the laws are vague, and unfortunately a lot of this is defined as precedent by the courts. Which is probably why patent lawyers are so expensive and why litigation is so common.
Sorry to keep harping on this, but nothing in either of the qouted portions of your post (the law or the treaty) indicate the *transfer* of any rights to the original invention to the inventor of an improvement to the original. The first merely states what can be patented (which includes improvements to process, machines, etc. -- whether or not the thing they improve is covered by a patent, BTW). The second (GATT excerpt) merely says that the patent must involve an inventive step, i.e. you can't just randomly throw things together, but have to have reasoned that the specific arrangement is better.
Since super frobinator with cheeze is a complete product that does all that a frobinator but better faster and with a significant feature addition ie inventive step, my patented super frobinator is a seperate 'in whole' patent. If I make a product that 'superizes' a frobinator that is an improvment. Whether you get the superfrobinator or the superizer is up to the USPTO...
NO! If the SuperFrobinator works the way the Frobinator works (minus the improvement), then the maker of SuperFrobinators will need to license the Frobinator technology from Frob Inc. or risk being sued. International Manufacturing Co. v. Landon, Inc. 336 F.2d 723 (9th Cir. 1964), is an example of a real, historical suit which exempifies this.
To avoid this, what typically happens is that Frob, Inc. and SuperFrob, Inc. cross-license each others stuff so that neither can sue the other (if the SuperFrob improvement is really substantial, Frob wants to enter into a licensing agreement because if they don't, they believe that the SuperFrob enhancements will lead to increased sales, and SuperFrob wants to because they can't sell their product otherwise; if the SuperFrob enhancements are more like the So-SoFrob, then Frob choose to simply keep selling their version and bleed SuperFrob or refuse to license to SuperFrob, in which case SuperFrob is screwed).
On the other hand, the separate Superizer from SuperFrob, Inc. (for use with Frobinator by Frob, Inc) is fine because it doesn't *contain* an embodiment of the original patented invention.
The whole purpose of patents is to grant exclusive use of an invention to a patent holder for a limited time; if the holder of a patent for an improvement got rights to the original, patents would be rendered meaningless because anyone could come along and make some minor improvement and the original patent holder's monopoly is abolished.
Slashdot Mirror
More at stake: the power to use technology you purchase in an way you see fit (allowing people to directly maximize their experience of several important individual liberties, not least of which is the freedom of speech).
What new laws will help reward artists: Gee, how about a tax credit for payment to artists and scientists?
Many kids will either own computers or work with them daily after high school. Some may want to go on to work in an explicitly computer-oriented career, which however doesn't require much specific post-secondary education (hardware repair in a small shop, for instance). To the same degree as auto-mechanics (and probably far more than say, wood shop), computer education in secondary schools is a valuable addition to the curriculum.
For primary education, heavily computer-centric instruction may be overkill. But at the high school level good arguments can be made for it.
Of course, it won't be too useful to students who just want their school to subsidize their bong-building activities, but that's what metal shop is for.
Is this good? NO! Is it common? In my experience, in the literature of our industry, and of the opinion of most of the programmers I personally know (which is a large sample), it is the *rule* rather than the exception, unfotunately.
I hope you get my drift. Do I personally believe that terrorists have infiltrated Microsoft and planted bugs in the code? Not likely. Is the scenario conceivable? Absolutely.
I.e., while the code was being modified during the GA process, the result is the most "fit" bit across all iterations, included statically in the final program.
For example:
Secondly, while I agree that it's unlikely that a terrorist would approach a 13-year old kid and say, "Hey, you should start excelling in Math and then attend college to get a CS degree so that 10 years from now you can go work at Microsoft for 4 years or so (enough to gain the confidence of your managers) and then start putting back doors and bugs in their OS," it's far more plausible that a terrorist would approach a already working programmer who's naive and idealistic -- and perhaps *already* working at and trusted by managers at Microsoft -- and say, "Hey, here's how you can really help your faith..."
That is why the meaning of words or phrases change over time. "Steep Learning Curve" means one thing technically in psychology, but another altogether in popular usage. Which version is correct depends on one's audience, not one's bias. If anything, it's likely that the meaning in popular usage will come to so dominate the original technical meaning, that psycologists will actually reverse the axes of the graph in the traditional examples and the old technical usage will become archaic.
The only issues that arise are when other people start to have some copyright interest in your product; then, in order to distribute it under proprietary terms, you must have their permission as well.
This is why, for instance, Sleepycat Software requires patch submitters to assign the copyright for the patch to Sleepycat before they apply the patch to the official Berkeley DB code. Something like that approach seems neccessary unless you want to end up maintaining two codebases: the GPL one and the proprietary one.
In the second, binary only scenario, the PITA of reverse engineering ensures that a much smaller # of both white and black hats will attach themselves to the problem, and J-Random-tinkerer can't contribute at all. You've made it less likely that some "bad" person will use the bug to hurt you, but you've also made it less likely that some "good" person will find it and help you avoid or minimize the damage from the "bad" person's discovery.
My point is, security is a gamble. You are always gambling that there are no bugs in your program that some bad person is going to discover and exploit to harm you. "Security through Obsurity" is the wishful thinking strategy; it seeks to minimize the # of people who find the bugs, and then hope that those people are only the "good guys." Open Source and full disclosure are the "hedge your bets" strategies. They seek to maximize the number of people who find the bug, hoping that at least *one* of them is one of the good guys.
Purely statistically speaking, which do you think is likely to be more successful?
88 + 167 = 255; In other words, the CNN Money article seems to be using the same figures you quoted previously (probably from a common source which is geared more toward PR than rigorous accounting).
It's probably more like:
88 + ((167 * x) - y) - 175 = n
where
- x = percentage of gross recouped by studios (after all, movie chains bid on pictures, and AFAIK, that's what the studio gets. So in the case of a US dog, where the foreign theaters say "based on poor performance in the US, I'll only pay you $50 million", and then the film grosses $167 million, it's not clear to me who gets the bulk of the money); and
- y = the cost of promoting the film overseas
I'd say that only the studio accountants really know whether n > 0.The Postman
Production costs: $80 million.
Domestic gross: $17.4 million
Linus lives in Silicon Valley, Alan in Swansea, Wales. That's one monster bus.
Smallpox is the largest known human virus. Its spead is primarily airborne. I think you may be confusing viruses and bacteria in this case.
Additionally, there is now evidence that Chimpanzee tissue culture -- Chimpanzee SIV being the supposed AIDS precursor virus -- was used to passage the seed virus to produce vaccine (something which every opponent of the OPV/AIDS hypothesis claims never happened, but of whom none can prove, given that no protocol for vaccine production has ever been produced for the Koprowski/Wistar vaccines in question), and circumstantial evidence that vaccine produced in this manner may actually have been used in field trials in humans.
So,
- The people who are claiming that the debate is over are the people with a compelling financial interest in squashing this line of investigation;
- The people who are claiming that the debate is over are ignoring evidence as they see fit, and are content not to ask for evidence (e.g., vaccine manufacture protocols and lab notes related actual vaccine production lots) which would shed a great deal of light on the issue;
- The proponents of the theory have no financial interest in the outcome of the debate(even if one could argue book sales, the books will sell even if the theory turns out to be false);
- ONLY THE OPV/AIDS theory accounts for the extremely high correlation of mass polio vaccinations and early AIDS infection sites, especially in remote, supposedly "isolated" areas in Eastern Congo, Western Rwanda/Burundi;
While I am pro-science, I am strongly pro-caution. We all want to rid our species of diseases, but we must carefully weigh the potential costs with the uncertain benefits and strive to reach the utmost level of certainty with regard to safety. Unfortunately, the rush to be the first to do something (and thus get the profits associated with commercial development, and secure future funding) can be highly detrimental to that goal.Unfortunately, due to the primitive state of advertising at the time, the revenue never really panned out, and the resort was abandoned.
On the one hand, I think that some people are just inherently critical thinkers -- I lump myself and many of my friends into that category. Whether this is innate, or due to experiences growing up (i.e., learning to distrust authority) is anyone's guess.
However, I think you may be right that context has something to do with it. We who spend lots of time on the internet have learned, as you point out, that the vast majority of information there is crap, of little value. Likewise, Joe Sixpack may be more critical of Gary Condit (because it's conventional wisdom that politicians lie, especially about sex) than "that feller my sister's friend's friend knows that had his kidney stolen during Mardis Gras."
Mind you, I don't believe that this is because people are "stupid." Our educational system, with its emphasis on rote memorization and deference to Authority, does a poor job of teaching people to evaluate information that claims to be "authoritative." For example, how many of those reading this post have received the "GOOD LORD, CONGRESS IS GOING TO TAX EMAIL!" chain spam from countless of their relatives and friends -- a hoax email which, due to the fact that it claims a non-existent Senator as the sponsor of a bill which breaks the pattern of identifiers for senate bills (i.e., doesn't start with SB) -- is easily identifiable as such?
Face it; most people lack the ability to evaluate the value of information on the Internet.
If your invention inherits the implementation of the other invention, you need to license the other invention. If your invention merely inherits the interface (i.e., it produces similar results to the original invention, but without using the components -- be they physical parts, or the steps in a patented process), then you don't have to license the original. In the latter case, it's unclear to me though whether this is a "patented improvement" per se, or just a new patent.
As you said, the laws are vague, and unfortunately a lot of this is defined as precedent by the courts. Which is probably why patent lawyers are so expensive and why litigation is so common.
To avoid this, what typically happens is that Frob, Inc. and SuperFrob, Inc. cross-license each others stuff so that neither can sue the other (if the SuperFrob improvement is really substantial, Frob wants to enter into a licensing agreement because if they don't, they believe that the SuperFrob enhancements will lead to increased sales, and SuperFrob wants to because they can't sell their product otherwise; if the SuperFrob enhancements are more like the So-SoFrob, then Frob choose to simply keep selling their version and bleed SuperFrob or refuse to license to SuperFrob, in which case SuperFrob is screwed).
On the other hand, the separate Superizer from SuperFrob, Inc. (for use with Frobinator by Frob, Inc) is fine because it doesn't *contain* an embodiment of the original patented invention.
The whole purpose of patents is to grant exclusive use of an invention to a patent holder for a limited time; if the holder of a patent for an improvement got rights to the original, patents would be rendered meaningless because anyone could come along and make some minor improvement and the original patent holder's monopoly is abolished.