I wish to suggest that this is the immediate solution. The complete solution involves a truckload of pissed-off users storming a POSIX committee meeting and bashing the committee members over the head with clue sticks.
Yeah, I can see that. Ok, I can see the Punchscan method would be safer. I think, though, you'd need to alter the rules a little and require a winner to have sufficient margin to exceed what could reasonably be done via fraud, with a run-off being declared if the margin is smaller than this.
In the end, I guess the two ideas are similar, except that mine shifts the danger spot to the individuals (which, as you point out, increases the vulnerability) whereas Punchscan has a more centralized attack vector which (in principle) is going to be harder to meaningfully attack and you can (as shown above) fix the rules to make such an attack largely ineffective.
I think the practical upshot, no matter what the preference of the individual, is that between the two of us, we have shown that there ARE ways of doing e-voting that is not exposed to fraud by election officials or voting machine officials, and that these are NOT going to be any more costly or difficult to install or maintain than the systems we know don't work.
Of course, the odds of getting the US Government to read Slashdot and actively respond to this thread by revising the rules for electronic voting are slighly under the odds of the atoms in your body spontaneously quantum tunneling to a far-distant planet of geeks who live in a city dedicated to your personal vices.
Isn't it funny that a bay right next door to agencies and military groups that want to dispose of "evidence" happens to be chock full of bugs that, well, dispose of "evidence"...
You are correct. The proposal I'm putting forward (where a voter needs to serially scan perhaps millions of votes) is - at least in principle - just too expensive for vote selling. At least with computers as they stand. The compute cycles required for any large-scale checking would require a significant piece of big iron, which means it won't be portable. To go round and check, using portable computers, would need an army of vote-checkers of a size comparable to the number of voters.
Adding random data would, of course, make any kind of inspection impossible, but it would also make auditing much harder and would make distributed monitoring by the electorate impossible.
Some would say that this is a good thing, but I'm of the mind that you have vote fraud on a large scale in every election because accountability is impossible. Any replacement system must, to be credible, produce a net gain in trust and that means increasing the ability to detect both vote injection and vote deletion. You cannot do that with 100% anonymity guarantee. What you CAN do is take a leaf from the book of encryption. Encryption doesn't make it impossible to decrypt, but it does make it too expensive to be worth it.
First, I disagree. E-voting as implemented cannot be trusted at all, but that does not mean E-voting in general cannot be.
Let us take the simple case where every transaction is placed in a transaction journal. It is never erased from that journal, no matter what. You can "delete" as many times as you like, each delete is itself just a transaction that is logged. You now have a fix for the above problem, and indeed for any other problem to do with summing things up, as each vote is independently stored in the log, NEVER a total. Totals are calculated by replaying the transaction journal, they are never what the machine directly uses or processes.
Ok, but what about vote injection? That can be done via the backdoor you mention. Easy fix for that. PKI. Each voter has a registration card, right? So store a public key on it and have a dictionary of private keys on a remote machine. The compiler can insert all the backdoors it likes, if there's no private key on the remote vote tally system, the vote cannot be decrypted and therefore cannot affect the total.
Ok, what about backdoors which allow a hacker to directly delete data from main memory? Easy fix for that too. Never rely on the voting machine to be a file store. Each time a vote is cast, have it transmitted. In fact, if you really want, have it multicast. Anyone who wants to count the number of cast votes can then join the multicast group and count votes. They can't see who voted, they can't see who was voted for (it's still encrypted), but they CAN tell if the total number of votes sent to the multicast group is equal to the number of votes counted plus the number of votes rejected.
So far, so good. Now, what about making sure that votes aren't fraudulantly rejected? Easy. The private keys, when generated, can be duplicated and a spare set placed in a neutral archive. If there is a legal challenge, the duplicate keys can be removed from the archive and (a) compared with the keys used by the tally computer, and (b) used to decrypt the votes gathered. It doesn't matter who gathers the encrypted votes, so you can use the encrypted votes gathered by UN monitors if you really insist. The UN can't tamper with them any more than anyone else. The most they can do is randomly delete them, but since everyone has a copy of the encrypted votes (including all parties), this would be instantly open to legal challenge itself.
What about privacy in voting? Well, when the key pair is generated, simply split the pair. If there is no connection in any part of the system, after generation, to link the public key to the private key, nobody at all can tell who cast which vote.
(Well, technically the voter can tell, as they have their public key and know what vote they cast, so can re-generate the vote, re-encrypt it, and look to see if a vote posted over the network matches the vote that was re-calculated. But nobody else could do this, and given the time overheads, this could never be used to check up on voters to see who they voted for. It could only be used by voters themselves to ensure their vote was in the system.)
You don't have a "perfect" system after doing all that, but it's damn close to perfect, in that it meets the paper trail requirements, it is virtually impossible to add or remove votes without detection, retains anonymity, and yet is completely open to audit by anyone who feels like putting in the time and effort. There isn't a paper ballot system on the planet that can compare with that degree of integrity checking, and the basic paper ballot is one of the best systems ever devised.
There may still be some videos on YouTube of a chemistry lecture series done by the late Dr. John Salthouse from the University of Manchester. He ran a series of lectures which demonstrated all kinds of ways to blow things up. Liquid oxygen on rich tea biscuits (a UK cookie) was one of his favourites. Igniting steel wool with a 9v battery was another. There might be something that could be used in a high school that would be impressive enough and still legal.
Great Egg Race projects. Easy to establish a problem in physics and engineering that students can solve that are hard enough to be challenging and which do something that is obviously interesting.
Astronomy. There's plenty of Open Source code for creating images for amateur astronomy. What there isn't (at least, as far as I know) is any software for doing visible light interferometry. Set up a cluster of telescopes (they're cheap enough) and get the students to link them together as a single interferometry array. Also gives them something to show on a website, as space pics are (as others have noted) always popular.
Chaos Theory. Always good for some fun. Lorenz Waterwheels can be built for real, they are not just theoretical. But can you take a specific mathematical model and translate it into the corresponding physical model?
Archaic physics. Also always good for some fun. It turns out to be quite difficult to build a simple A-Frame and pulley system to lift 50 tonne stones. Most attempts have failed, although the correct technique must be close to what is being tried. Can the students fix the bugs before they get squished like bugs?
There is currently no physical or chemical model which can tell you what colour a pottery glaze will produce when fired, either in a modern kiln (which is oxidizing) or in an archaic kiln (which is reducing). By simulating the physics of what is taking place and the physics of optics, see if it is possible to produce some sort of a model, even if not universal or terribly accurate, that is better than the try-and-see method that potters currently use.
Semiconductors. We know that it's possible to make graphene with nothing more than pencil lead and lors of scotch tape. But what, exactly, ARE the semiconductor properties of graphene? How do these change with temperature? (Liquid nitrogen isn't hard to get hold of.)
High-Energy Physics. Once you're done with graphene, we know that scotch tape is an excellent source of X-Rays - when in a partial vaccuum. What energies can be produced? Do these change with a change in the level of vaccuum? What other factors might change the energies?
We all know about building a paper structure that can contain a raw egg, be dropped off a building, and have the egg intact at the end. But paper isn't terribly rigid. How far can you scale such models? Which models scale the best? How many eggs can you pack in in any given model?
Fluid Dynamics. There are well over 1,500 "commonly used" aerofoils published. It's not hard to make these and attach them to springs to measure forces, and if the Wright Brothers could make a wind tunnel using a bicycle and fan blades, so can a student. Compare the observations with what you'd expect based on different mathematical models (eg: the Bernouli Effect).
I can agree with some of the points you raise, and yes, it's a lot more complex than "just" the scriptwriter - I was oversimplifying things there a bit, I'll admit.
I'll also agree that what is stocked is a function of demand (which is why I think Linux advocates would be better gauging general interest by changes in floorspace and shelfspace than by changes in the number of downloads from a given FTP site).
And, yes, it's very very hard to deconstruct a good movie and allocate a share of the responsibility to the scriptwriter. (It's much easier to deconstruct bad movies, as there is usually some indication of what the bad part was. Rubbish stands out far more in the minds of an audience, so it's easy to see what it was that was actually rubbish.)
However, I would argue that acting schools tend to be fairly uniform in what they teach and how they teach, at any given time, although the methods have changed dramatically over the decades. I am not convinced that actors within a given generation are going to have nearly the same variability as the writers within a single generation. The biggest selling-points with actors seem to be name and looks. I can't think of any case of an actor who got tapped for a part because of differentiation from the norm.
The money being invested is a factor, certainly, although from what I understand, the original Star Wars had a budget not dramatically larger than those of higher-end independent movies. The documentary penguin movie that out-performed The Fabulous Four also didn't have a huge budget.
Besides, it's common knowledge the budgets are all forged to ensure that no movie can ever make a profit. Estimates I've seen from commercial independent movie websites suggest the big studios inflate budgets between x10 and x100 to rig the numbers, mostly for tax reasons but also because newspapers are more likely to cover a movie with a budget that dwarfs the whole of Formula One racing than they are to cover a movie that was produced on the cheap.
So although I'm willing to accept these other factors are present and may indeed by very significant, I'm not convinced they drown out the contribution of the writer.
Not sure about the density Hales-Jewett theorum, but the method used seems simple enough. The problem he is trying to tackle is not one of determining a proof for the problem, but rather establishing that there either exists a proof for the problem given a set of parameters or does not exist a proof given those same parameters.
To me, it is this problem which seems to be the more useful, at least in general. Basically, if it is possible, through herustics (which is essentially all you're doing by hunting through possibilities thrown out by the combinatorial logic), to determine if a solution must exist (or must not exist) to a given problem given constraints, we can use the same approach to solve quite a number of "there exists"-type problems herustically.
In the case of Linux, this would suggest that if you can define tight enough constraints on a given module (the equivalent of the constraints imposed in the original problem), it should be possible to prove whether or not an arc exists through that module which would violate the constraints without having to find that arc.
If an arc exists that would violate the constraint, that would be the same as finding a "good reason" why the constraints will not work in the original problem addressed in TFA.
This would seem to offer an approach to verification of even fairly complex software, without having to clear all of the hurdles raised by formal software proofs. If you don't have to find the buggy arc, only show that such an arc must exist, then you can use this to identify areas in the code which need fixing.
Since you can bisect any code and then test each of the subcomponents, you can narrow down which blocks of code are faulty without having to be able to prove why, how or where. That part can then be addressed manually.
This would semi-automate at least some elements of bug-hunting and quality control. Even if you could only use the approach in a subset of kernel modules, due to the complexities of interactions, I see no reason why you couldn't use this to perform far more rigorous checks than things like the Stanford Checker or Klokwork are currently capable of.
Short-term financial success does indeed depend on the industry as a whole. Independence Day had no script at all, just a bunch of explosions and some loonies running around in small circles, but made a fortune. For a few days. I seriously doubt it has shown any kind of sustained earning power.
Long-term financial success depends on critical success. People watch classics, and classics are usually not the ones topping the weekly earnings chart on their release. When movies are re-run on TV, the residuals paid to the studio go up with age (according to execs at scifi conventions anyways), so anything that gets re-runs stretching into the decades is going to have sizeable earning power, at least until copyright expires.
The implication is that it doesn't matter if 2001 didn't earn as much in its first week as Jurassic Park, as 2001 has likely been shown more often and been shown for longer. The better the script is, the longer it will be profitable for the stations TO re-run it, and the more profitable it will become for the studio. A naff script will only last for as long as the film is in the theatres, which might only be a couple of weeks, three or four at most.
So, yes, you're absolutely right if you're going by what is on the balance sheet. Balance sheets do not show projected earnings, except when the accountant is trying to pull off some fraud or other, and certainly do not show accumulated earnings. They're snapshots and I don't care how popular a classic is, the snapshots are ghastly. But it's just a snapshot. The junk pack-em-in movies only have a couple of snapshots, the classics have a stack of albums ceiling-high.
I can't see any obvious rough edges. We all know Santa won't/can't visit a house if anyone's awake to do the observing, that's one of the fundamental rules.
All objects accelerate towards the Earth at the same rate, regardless of mass, so no. Besides, on that scale, the forces between molecules is vastly greater than the force of gravity, which is why Brownian Motion is a realistic way to model a gas.
I would agree - to a degree. To me, an AI should be as smart as possible (even if superhumanly so - if I wanted a human opponent, I'd go to a gaming club), but should do so on no more information than a human player would have. Thus, you should not have one side play in a "fog of war" and the other be given a full-information scenario. That doesn't cut it.
But within the constraint of equally limited data, I have no objection to the computer throwing every clock-cycle it has into trying to beat me. I'd prefer it. Game AIs are frequently dumb to the point of being pointless.
One wargame I used to play was "Crusade in Europe". I found out that if you bombed the enemy supply depots and just sat on the beaches of Normandy, the AI's forces would all starve to death. I successfully won World War 2 from the D-Day landings onwards with under 500 casualties. I wish to argue that this should be impossible, no matter HOW good the human player is.
Sure, players want to win. That's natural. But they should win because they're good enough to win, not because the AI lets them, even if the AI is sneaky enough to not make it obvious that it's letting them win. Games should be hard. It took me almost a year to reach the top rank in BBC Elite. Had the AI been half-way competent, it should have taken me longer. Games that are completed and disposed of in a fortnight aren't worth the money to buy or the effort to write.
All atoms move randomly. Under normal circumstances, they move in all different directions. Statistically, if you stand on top of a building for long enough, they will all move the same way, out into the open skies. (And then back to random motion as you plummet towards the ground. Who said QM doesn't have a sense of humour?)
Regardless of what any contract says, regardless of who actually owes what, screenplay writers are the major breadwinners yet get paid virtually nothing for their efforts. Nobody got rich writing scripts, but many many rich actors and movie moguls got rich from bloody good stories.
Now, onto the crux of what he says. It is well-known that money brought in through lawsuits, etc, via the MPAA and RIAA have not been forwarded to artists. It is also well-known that artists repeatedly sue managers, producers and studios for payment of royalties. Is it too hard to imagine the studios rip off those who are respected and heard even less?
The totals are probably exaggerated a little. A Star Trek FAQ from the 1990s suggested the annual turnover of Star Trek merchandise was around 60 million dollars. Recent FAQs don't show any estimate and deny it's possible to calculate one, so this is the only figure I can really go on. It simply isn't possible for a single episode (minus residuals owed to everyone else involved) to be worth hundreds of millions of dollars, even if we assume the FAQ figure to be about right. Tens of millions, divided amongst everyone, for the entire time since original screening - that sounds more likely.
Given the number of people involved wasn't many, I could see that he should have made somewhere in the low single-digit millions or upper three-digit thousands off a single script at this point. If he has made less than this, he has every right to feel like the studio is ripping him off.
Of course, legally, all that matters is what the contract says. If the contract says he should be paid X amount and he has been paid less than that (a common enough experience with artists, so why not writers?), then he has not just a moral argument but a legal argument.
Those who accuse him of kicking up a fuss over nothing should remember that the studios ARE rip-off merchants, and ARE making a great deal of money off Star Trek. There isn't the slightest possibility all of the money Paramount is making is legal. Maybe most of it is, but don't expect me to believe they're being honest for the first time in their lives over one of their biggest money-spinners. Their lawyers are bigger and their accountants are sharper. If there's a way for them to have hidden income, you can be certain they have.
It is a general truism that if you're going to make a major discovery, it's going to be in your 20s or early 30s, no later. Most doctors probably aren't going to need a mental capacity great enough to make earth-shattering discoveries, but given the mind slows down dramatically in the 40s and beyond, I do believe doctors will decline after that point.
I would suggest, though, that this research means that people making critical decisions should be as close to 27 as possible. The research showing a strong decline at 40 suggests this should be a maximum age for any mission-critical jobs (like being President or being a Supreme Court Justice). Experience isn't much use if you can't remember it or figure out how it relates.
Slashdot Mirror
I wish to suggest that this is the immediate solution. The complete solution involves a truckload of pissed-off users storming a POSIX committee meeting and bashing the committee members over the head with clue sticks.
Yeah, I can see that. Ok, I can see the Punchscan method would be safer. I think, though, you'd need to alter the rules a little and require a winner to have sufficient margin to exceed what could reasonably be done via fraud, with a run-off being declared if the margin is smaller than this.
In the end, I guess the two ideas are similar, except that mine shifts the danger spot to the individuals (which, as you point out, increases the vulnerability) whereas Punchscan has a more centralized attack vector which (in principle) is going to be harder to meaningfully attack and you can (as shown above) fix the rules to make such an attack largely ineffective.
I think the practical upshot, no matter what the preference of the individual, is that between the two of us, we have shown that there ARE ways of doing e-voting that is not exposed to fraud by election officials or voting machine officials, and that these are NOT going to be any more costly or difficult to install or maintain than the systems we know don't work.
Of course, the odds of getting the US Government to read Slashdot and actively respond to this thread by revising the rules for electronic voting are slighly under the odds of the atoms in your body spontaneously quantum tunneling to a far-distant planet of geeks who live in a city dedicated to your personal vices.
Isn't it funny that a bay right next door to agencies and military groups that want to dispose of "evidence" happens to be chock full of bugs that, well, dispose of "evidence"...
You are correct. The proposal I'm putting forward (where a voter needs to serially scan perhaps millions of votes) is - at least in principle - just too expensive for vote selling. At least with computers as they stand. The compute cycles required for any large-scale checking would require a significant piece of big iron, which means it won't be portable. To go round and check, using portable computers, would need an army of vote-checkers of a size comparable to the number of voters.
Adding random data would, of course, make any kind of inspection impossible, but it would also make auditing much harder and would make distributed monitoring by the electorate impossible.
Some would say that this is a good thing, but I'm of the mind that you have vote fraud on a large scale in every election because accountability is impossible. Any replacement system must, to be credible, produce a net gain in trust and that means increasing the ability to detect both vote injection and vote deletion. You cannot do that with 100% anonymity guarantee. What you CAN do is take a leaf from the book of encryption. Encryption doesn't make it impossible to decrypt, but it does make it too expensive to be worth it.
First, I disagree. E-voting as implemented cannot be trusted at all, but that does not mean E-voting in general cannot be.
Let us take the simple case where every transaction is placed in a transaction journal. It is never erased from that journal, no matter what. You can "delete" as many times as you like, each delete is itself just a transaction that is logged. You now have a fix for the above problem, and indeed for any other problem to do with summing things up, as each vote is independently stored in the log, NEVER a total. Totals are calculated by replaying the transaction journal, they are never what the machine directly uses or processes.
Ok, but what about vote injection? That can be done via the backdoor you mention. Easy fix for that. PKI. Each voter has a registration card, right? So store a public key on it and have a dictionary of private keys on a remote machine. The compiler can insert all the backdoors it likes, if there's no private key on the remote vote tally system, the vote cannot be decrypted and therefore cannot affect the total.
Ok, what about backdoors which allow a hacker to directly delete data from main memory? Easy fix for that too. Never rely on the voting machine to be a file store. Each time a vote is cast, have it transmitted. In fact, if you really want, have it multicast. Anyone who wants to count the number of cast votes can then join the multicast group and count votes. They can't see who voted, they can't see who was voted for (it's still encrypted), but they CAN tell if the total number of votes sent to the multicast group is equal to the number of votes counted plus the number of votes rejected.
So far, so good. Now, what about making sure that votes aren't fraudulantly rejected? Easy. The private keys, when generated, can be duplicated and a spare set placed in a neutral archive. If there is a legal challenge, the duplicate keys can be removed from the archive and (a) compared with the keys used by the tally computer, and (b) used to decrypt the votes gathered. It doesn't matter who gathers the encrypted votes, so you can use the encrypted votes gathered by UN monitors if you really insist. The UN can't tamper with them any more than anyone else. The most they can do is randomly delete them, but since everyone has a copy of the encrypted votes (including all parties), this would be instantly open to legal challenge itself.
What about privacy in voting? Well, when the key pair is generated, simply split the pair. If there is no connection in any part of the system, after generation, to link the public key to the private key, nobody at all can tell who cast which vote.
(Well, technically the voter can tell, as they have their public key and know what vote they cast, so can re-generate the vote, re-encrypt it, and look to see if a vote posted over the network matches the vote that was re-calculated. But nobody else could do this, and given the time overheads, this could never be used to check up on voters to see who they voted for. It could only be used by voters themselves to ensure their vote was in the system.)
You don't have a "perfect" system after doing all that, but it's damn close to perfect, in that it meets the paper trail requirements, it is virtually impossible to add or remove votes without detection, retains anonymity, and yet is completely open to audit by anyone who feels like putting in the time and effort. There isn't a paper ballot system on the planet that can compare with that degree of integrity checking, and the basic paper ballot is one of the best systems ever devised.
I believe the correct quote is "We need more cowbell".
Oh, that's easy. Always have the question be: "Why is this an interesting experiment?"
There may still be some videos on YouTube of a chemistry lecture series done by the late Dr. John Salthouse from the University of Manchester. He ran a series of lectures which demonstrated all kinds of ways to blow things up. Liquid oxygen on rich tea biscuits (a UK cookie) was one of his favourites. Igniting steel wool with a 9v battery was another. There might be something that could be used in a high school that would be impressive enough and still legal.
I can agree with some of the points you raise, and yes, it's a lot more complex than "just" the scriptwriter - I was oversimplifying things there a bit, I'll admit.
I'll also agree that what is stocked is a function of demand (which is why I think Linux advocates would be better gauging general interest by changes in floorspace and shelfspace than by changes in the number of downloads from a given FTP site).
And, yes, it's very very hard to deconstruct a good movie and allocate a share of the responsibility to the scriptwriter. (It's much easier to deconstruct bad movies, as there is usually some indication of what the bad part was. Rubbish stands out far more in the minds of an audience, so it's easy to see what it was that was actually rubbish.)
However, I would argue that acting schools tend to be fairly uniform in what they teach and how they teach, at any given time, although the methods have changed dramatically over the decades. I am not convinced that actors within a given generation are going to have nearly the same variability as the writers within a single generation. The biggest selling-points with actors seem to be name and looks. I can't think of any case of an actor who got tapped for a part because of differentiation from the norm.
The money being invested is a factor, certainly, although from what I understand, the original Star Wars had a budget not dramatically larger than those of higher-end independent movies. The documentary penguin movie that out-performed The Fabulous Four also didn't have a huge budget.
Besides, it's common knowledge the budgets are all forged to ensure that no movie can ever make a profit. Estimates I've seen from commercial independent movie websites suggest the big studios inflate budgets between x10 and x100 to rig the numbers, mostly for tax reasons but also because newspapers are more likely to cover a movie with a budget that dwarfs the whole of Formula One racing than they are to cover a movie that was produced on the cheap.
So although I'm willing to accept these other factors are present and may indeed by very significant, I'm not convinced they drown out the contribution of the writer.
Not sure about the density Hales-Jewett theorum, but the method used seems simple enough. The problem he is trying to tackle is not one of determining a proof for the problem, but rather establishing that there either exists a proof for the problem given a set of parameters or does not exist a proof given those same parameters.
To me, it is this problem which seems to be the more useful, at least in general. Basically, if it is possible, through herustics (which is essentially all you're doing by hunting through possibilities thrown out by the combinatorial logic), to determine if a solution must exist (or must not exist) to a given problem given constraints, we can use the same approach to solve quite a number of "there exists"-type problems herustically.
In the case of Linux, this would suggest that if you can define tight enough constraints on a given module (the equivalent of the constraints imposed in the original problem), it should be possible to prove whether or not an arc exists through that module which would violate the constraints without having to find that arc .
If an arc exists that would violate the constraint, that would be the same as finding a "good reason" why the constraints will not work in the original problem addressed in TFA.
This would seem to offer an approach to verification of even fairly complex software, without having to clear all of the hurdles raised by formal software proofs. If you don't have to find the buggy arc, only show that such an arc must exist, then you can use this to identify areas in the code which need fixing.
Since you can bisect any code and then test each of the subcomponents, you can narrow down which blocks of code are faulty without having to be able to prove why, how or where. That part can then be addressed manually.
This would semi-automate at least some elements of bug-hunting and quality control. Even if you could only use the approach in a subset of kernel modules, due to the complexities of interactions, I see no reason why you couldn't use this to perform far more rigorous checks than things like the Stanford Checker or Klokwork are currently capable of.
Short-term financial success does indeed depend on the industry as a whole. Independence Day had no script at all, just a bunch of explosions and some loonies running around in small circles, but made a fortune. For a few days. I seriously doubt it has shown any kind of sustained earning power.
Long-term financial success depends on critical success. People watch classics, and classics are usually not the ones topping the weekly earnings chart on their release. When movies are re-run on TV, the residuals paid to the studio go up with age (according to execs at scifi conventions anyways), so anything that gets re-runs stretching into the decades is going to have sizeable earning power, at least until copyright expires.
The implication is that it doesn't matter if 2001 didn't earn as much in its first week as Jurassic Park, as 2001 has likely been shown more often and been shown for longer. The better the script is, the longer it will be profitable for the stations TO re-run it, and the more profitable it will become for the studio. A naff script will only last for as long as the film is in the theatres, which might only be a couple of weeks, three or four at most.
So, yes, you're absolutely right if you're going by what is on the balance sheet. Balance sheets do not show projected earnings, except when the accountant is trying to pull off some fraud or other, and certainly do not show accumulated earnings. They're snapshots and I don't care how popular a classic is, the snapshots are ghastly. But it's just a snapshot. The junk pack-em-in movies only have a couple of snapshots, the classics have a stack of albums ceiling-high.
I dunno. JK Rowling isn't bad, even in her old age.
I can't see any obvious rough edges. We all know Santa won't/can't visit a house if anyone's awake to do the observing, that's one of the fundamental rules.
All objects accelerate towards the Earth at the same rate, regardless of mass, so no. Besides, on that scale, the forces between molecules is vastly greater than the force of gravity, which is why Brownian Motion is a realistic way to model a gas.
I would agree - to a degree. To me, an AI should be as smart as possible (even if superhumanly so - if I wanted a human opponent, I'd go to a gaming club), but should do so on no more information than a human player would have. Thus, you should not have one side play in a "fog of war" and the other be given a full-information scenario. That doesn't cut it.
But within the constraint of equally limited data, I have no objection to the computer throwing every clock-cycle it has into trying to beat me. I'd prefer it. Game AIs are frequently dumb to the point of being pointless.
One wargame I used to play was "Crusade in Europe". I found out that if you bombed the enemy supply depots and just sat on the beaches of Normandy, the AI's forces would all starve to death. I successfully won World War 2 from the D-Day landings onwards with under 500 casualties. I wish to argue that this should be impossible, no matter HOW good the human player is.
Sure, players want to win. That's natural. But they should win because they're good enough to win, not because the AI lets them, even if the AI is sneaky enough to not make it obvious that it's letting them win. Games should be hard. It took me almost a year to reach the top rank in BBC Elite. Had the AI been half-way competent, it should have taken me longer. Games that are completed and disposed of in a fortnight aren't worth the money to buy or the effort to write.
Well, they most be real. They've not been declared integer.
All atoms move randomly. Under normal circumstances, they move in all different directions. Statistically, if you stand on top of a building for long enough, they will all move the same way, out into the open skies. (And then back to random motion as you plummet towards the ground. Who said QM doesn't have a sense of humour?)
Look, this should be very easy. Think the Star Trek prequel. Isn't it obvious Paramount should be hung out to dry as punishment?
Regardless of what any contract says, regardless of who actually owes what, screenplay writers are the major breadwinners yet get paid virtually nothing for their efforts. Nobody got rich writing scripts, but many many rich actors and movie moguls got rich from bloody good stories.
Now, onto the crux of what he says. It is well-known that money brought in through lawsuits, etc, via the MPAA and RIAA have not been forwarded to artists. It is also well-known that artists repeatedly sue managers, producers and studios for payment of royalties. Is it too hard to imagine the studios rip off those who are respected and heard even less?
The totals are probably exaggerated a little. A Star Trek FAQ from the 1990s suggested the annual turnover of Star Trek merchandise was around 60 million dollars. Recent FAQs don't show any estimate and deny it's possible to calculate one, so this is the only figure I can really go on. It simply isn't possible for a single episode (minus residuals owed to everyone else involved) to be worth hundreds of millions of dollars, even if we assume the FAQ figure to be about right. Tens of millions, divided amongst everyone, for the entire time since original screening - that sounds more likely.
Given the number of people involved wasn't many, I could see that he should have made somewhere in the low single-digit millions or upper three-digit thousands off a single script at this point. If he has made less than this, he has every right to feel like the studio is ripping him off.
Of course, legally, all that matters is what the contract says. If the contract says he should be paid X amount and he has been paid less than that (a common enough experience with artists, so why not writers?), then he has not just a moral argument but a legal argument.
Those who accuse him of kicking up a fuss over nothing should remember that the studios ARE rip-off merchants, and ARE making a great deal of money off Star Trek. There isn't the slightest possibility all of the money Paramount is making is legal. Maybe most of it is, but don't expect me to believe they're being honest for the first time in their lives over one of their biggest money-spinners. Their lawyers are bigger and their accountants are sharper. If there's a way for them to have hidden income, you can be certain they have.
Pickling is an excellent way to preserve. Everyone knows that.
I prefer salmon sushi. It's less effort and there is about the same amount of scientific evidence that it's effective.
It is a general truism that if you're going to make a major discovery, it's going to be in your 20s or early 30s, no later. Most doctors probably aren't going to need a mental capacity great enough to make earth-shattering discoveries, but given the mind slows down dramatically in the 40s and beyond, I do believe doctors will decline after that point.
I would suggest, though, that this research means that people making critical decisions should be as close to 27 as possible. The research showing a strong decline at 40 suggests this should be a maximum age for any mission-critical jobs (like being President or being a Supreme Court Justice). Experience isn't much use if you can't remember it or figure out how it relates.
*blink*
You mean, you took my post seriously, even after suggesting Easter was based on throwing eggs at a cartoon character?
I think there is indeed a whoosh here, but I don't think it's anything going over my head.
And politicians. Don't forget them.