Actually, from the clips I have seen of the movie, Stark clearly has genuine human level (if not super-human) AI available.
He has a design support system consisting of virtual reality and robotics controlled by a sophisticated AI system. The AI system is capable of determining what its master intends to do, projecting its probable results, feeling concern, and expressing and explaining its disagreement.
That's beyond what a typical computer can do even in Star Trek.
I think the computational problem of translation is one that is amenable to the kinds of advances we've seen in the past few decades in computer chess. The big difference is that chess is a self-contained world, whereas language refers to things in the outside world that the speaker assumes the hearer knows in common with him. Even people do a pretty crappy job of understanding when people are coming from different backgrounds or even starting from different assumptions.
The experiential aspect of communication is probably going to be the greatest hurdle in high quality translation. "Real time" is must a matter of computational bandwidth.
Well, I'll debate the science merits of Harry Potter.
Magic, in Harry Potter, is depicted as much more scientific the science is in most sci-fi. It is more technological than mechanical engineering is in Iron Man.
It is true that magic, in Harry Potter, can only be done by some people. However if you are one of those people, it operates in a repeatable way. In fact, it is not just repeatable in a hocus pocus way, but in a way that allows people to abstract general laws and principles for its operation. Magical techniques and knowledge are improved both by practical experimentation and theoretical exploration. There are even scholarly journals in specialized magical sub-discplines ("Transfiguration Today", and "Challenges in Charming").
In fact in several cases these magical disciplines are actually real world cases of obsolete sciences that did not keep up with the empirical advances of science as a whole. Newton was an alchemist, after all, and numerology ("arithmancy") and astrology where related Hermetic proto-sciences much studied by Newton's contemporaries. Medicine barely squeaked over the line in the real world, leaving chiropractic just on the other side. Osteopathy was the last bit over the fence, and medicine is absorbing it into normal "allopathic" practice.
In Harry Potter's world economic competition spurs technological innovation in broomstick design. Research is discovering new treatments for ancient diseases (lycanthropy) and new applications for materials (dragon's blood). The government itself has advanced research facilities for advanced research into topics like the nature of time.
Education seems to follow this pattern: students are introduced to a theoretical concept, then they are given a practical demonstration, followed by a lab exercise. If they fail the lab exercise, they are given additional practice and theory assignments, culminating in an essay which demonstrates their grasp of the principles being taught. After a course of many such cycles, they are given an exam which determines their competency to study more advanced topics. In many ways, the educational practices of Hogwarts would be exemplary if it were an engineering college.
In most stories, magic is a kind of detritus of a past Golden Age; the knowledge of that age has been shattered and its pieces are slowly disappearing out of the world. Lucky (or sometimes very unlucky) individuals come into possession of bits of that knowledge, often in the form of ancient and hard-decipher scrolls or mouldy old grimoires. It is notable that Hogwarts texts, while at times a bit -- eccentric, are generally freshly printed and (sometimes) represent the efforts of reasonably contemporary scholars.
In short, the world of Harry Potter is most strongly characterized by progress in theoretical knowledge (science) and practical applications (technology). The one difference is that the minds of individual magic practitioners are interfaced in ways that are more direct than trough the the nerve endings in their bodies.
Well, not really. No way you can make a suit that flies that way for more than a few meters without anti-gravity. Where are the fuel tanks? And if a real person was inside of it, doing the things it supposedly does, you'd end up with a can of Campell's Hearty Tomato Stew with Real Meat Chunks.
And don't get me started on the issue of batteries. The existence of a battery system as good as Iron Man's is inconsistent with the world as we know it today.
What is different about the story is that it is located in a world that is more recognizably our own. The Batman movies are more blatantly archetypal. Gotham City is closer to Middle Earth than it is to our world. It isn't just that the story features real place names like "Vietnam" or "Afghanistan" (which in the story don't actually much resemble the real countries of the same name). It's that Tony Stark lives in a world where most people get along, and few like him do very well, by not rocking the boat, not questioning the status quo, and acting as if the aftermath of their actions were somebody else's job to clean up.
What gives the Iron Man story a different twist on most comic book stories is that if you strip away Tony Stark's engineering genius, he isn't really all that different from any other American; he just has the money to do indulge himself on a larger scale.
Well, consider this was 1964, it might have been a computed goto...
In any case, this was not a language intended for software developers (like Algol). Nor was it a language intended for scientists (like Fortran). It was intended for CS students. Goto is desirable for demonstrating simple models of computation.
Dartmouth Basic had if/then, gosub, and for/next loops, which were much less gawdawful than Fortran's Do loops. It had fewer than a dozen functions, but they were well chosen to give students the ability to do interesting, non-trivial stuff.
I started programming maybe a dozen years after Dartmouth BASIC was introduced, and many of the interpreted language options available were not nearly so clean and well thought out.
Of course, that in part was the downfall of BASIC; it was well enough designed for its purpose that its impact was much larger than its target audience. Extending the language and libraries to such a wide variety of practical uses diluted the virtues of its original design. It was no longer a minimalistic language and runtime environment for students to learn and demonstrate concepts and for academics to do computations in. Its descendants became the patronizing choice when you targeted people who you assumed unable to learn anything better.
I've used NTFS3G under fuse for mounting Windows partitions with virtual machines stored on them.
It's remarkably non-sucky, considering it is a user space reverse engineering of a proprietary FS. The maintainers haven't really even attempted to optimize the thing, and they were rewarded for their focus on correctness with -- not exactly outstanding performance, but surprisingly good performance under the circumstances.
The main drawback I can see is that it uses a lot of CPU for large file operations like copying multi-gb files.
Still, hats off to the developers for an exceptionally useful bit of software. I haven't found a single serious glitch in sharing NTFS partitions with Vista.
I spent some time evaluating file systems for hosting virtual machine disks, and found JFS had a slight edge in XFS in consistency of performance, although XFS was perhaps a bit faster on average. To tell you the truth, even ext3 was quite usable. From a subjective standpoint the spread was not all that dramatic with the workloads I have.
In some ways, I remain very impressed with ext2/ext3, which was never designed to be the last thing in advanced file system. It was designed in 1993 to be a solid, all around performer and it remains so fifteen years later, which is amazing when you think about it. People say its performance is terrible, and that may well be true if you compare it to a file system with specialized strengths in the area of those strengths.
Still, I settled on JFS, which is probably the most rock-solid file system I've ever used. XFS has a reputation for losing data in power loss situations, which I haven't personally encountered. I have encountered data corruption on Reiser under similar conditions. For that reason I'd consider XFS or Reiser mainly if I thought I had a particular need their areas of exceptional competence.
I'm happy with jfs for all around use. It seems to be a very reasonable default choice, because it doesn't seem to have any glaring weaknesses. It performs solidly in a wide variety of situations, it handles heavy disk loads with low CPU usage.
Not a problem. Just let the user print bogus receipts.
At any point in the voting process, you can print a receipt. That receipt contains a cryptographic checksum of the voting, including the fact that the receipt was preliminary. Upon the actual casting of the vote a genuine receipt is printed.
Subsequently, if you want to verify a receipt, you need a court order to recover the hardware crypto key, which has been removed from the machine at the close of voting and placed at a secure escrow site.
True, but you don't have the scenario where the torturer is standing next to the prisoner with a cattle prod in his hand as the prisoner gets his email. The prisoner doesn't know whether it's actually his lawyer on the other end, and not an agent who has stolen the lawyer's key.
Prisoners in terror cases probably don't even have the right to send and receive unmonitored email. Prisoners don't have the right to make unmonitored calls.
Finally, you can't see that your client is haggard from sleep deprivation and rail thin from being starved and bruised from being beaten. The prisoner doesn't get the same reassurance that he hasn't disappeared into an anonymous void from exchanging emails.
I'm not saying that cryptography doesn't have some application in these situations -- although perhaps more for the coordination of the defense than for communicating with prisoners. I'm saying there is no substitute for being there.
Of course there was evidence. The evidence wouldn't have been enough to convict, but Reiser made tactical mistake of trying to bullshit his way out of it instead of leaving that to the closing arguments. All he had to do was leave the door open to reasonable doubt.
In the end, it is true that Reiser was convicted because he was arrogant. But while his arrogant demeanor probably didn't endear him to the jury, it was what his arrogance lead him to do that convicted him. His own testimony was evidence.
The system would be grossly unfair if you had to prove your innocence, but you don't. You don't have to have a theory that covers every thing. You just have to point out the holes in the prosecution's theories. Anybody who's watched reruns of Columbo knows that if you ad lib long enough, you get caught.
So he took the seat out of his car because he was sleeping in it. Didn't he notice there was an inch of water in the car? He didn't remember sleeping in a puddle of water. That's what got him convicted, he ran out of explanations and said something really, really stupid. What happens in the jury is that with twelve ordinary people, somebody is going to say, "well, maybe he planed to sleep in the car but didn't because it was full of water." And there you've got reasonable doubt. If you try to remove all doubt, you discredit the universe of possiblities under which you might be innocent.
In the end, Reiser's tragic flaw was that he was too arrogant to suppress his foolish impulses. Only an fool kills his wife to fix his problems. Even so, Reiser was smart enough to get away with it, but he was too arrogant to keep his mouth shut. If he had the kind of intelligence it took to restrain his mouth, he'd never have been in the situation where it needed restraining.
Um, no, we don't. For one thing, I don't see "the non-support of non-free software as a necessary condition for having free software". OTOH, I do see expenditure of resources developing software for a non-Free platform as a choice that reduces the available resources for improving the quality of the software stack on the Free platform.
But this is demonstrably false as a general statement. Given that you must pay for both hardware and software, a software strategy which allows you to reduce your average cost spent on hardware could conceivably increase the total resources you have available, thus enabling you to spend more money supporting free software in the future.
I'm not saying it is necessarily so in this case, I couldn't without looking at the project's planning documents. I'm just saying that the way you budget for a product isn't quite like a household budget in which for the most part your discretionary costs are operational, not capital in nature.
you see it as a fixed fact that the major cost of the systems will be the hardware cost, whereas I see that as being true if and only if the software and content is Free.
Well, your argument would make sense if the plan was to abandon the idea of an entirely free software stack. Is that what Negroponte is advocating? I don't believe so.
Even free (gratis) software and content from a commercial vendor is likely to become a major cost down the road in maintenance and upgrades
Which is only possible to the degree that users depend on Windows only features. Since the shell is Sugar, the real question is whether users will become dependent on certain Windows only programs, and whether F/OSS replacements will not be available because of the monopoly effects of MS.
But I don't think that's very likely. Given the device's market, users will actually find the selection of Linux software much wider. This is just one of those things that customers think they absolutely have to have. After they have the things, find them useful, then they'll see their conception of the challenges was upside down. They'll ask, is there a chemistry instruction program for Windows? Yes, it cost $$$, but there's a free Linux one.
Sooner or later people will just ditch Windows and run Linux, because running Windows on this thing will obviously be stupid -- for the target audience. Heck, the kids can install Linux to get at software nobody will buy for them. They probably will. But they won't be able to if they don't have the hardware, and they won't have the hardware unless certain parties have their Windows security blanket.
Linux itself owes its existence to the desire to have something better to run on your cheap commodity hardware. Free software for the third world is dependent on there being hardware which means the conditions and means of third world people.
I engineer a 150hp locomotive. I can pull your boat, your car, and maybe even your house.
Not on route I95, you can't. In any case, a 150hp locomotive? What is that, a kiddy ride? I thought locomotives were in the 2000 - 6000 hp range, with self-propelled railcars being in the 500-1000 range. You must be measuring horsepower by some different method. I know that steam engines can be rated by boiler horsepower but be capable of delivering much higher net horsepower over short distances, by a factor of five or more. The old Stanley Steamers with a 20HP rating could actually deliver 125HP for a while, before boiler pressure started to drop.
You said very specifically that if we spent more money on space exploration, it would be "a lot sexier", which in this context I presume to mean attractive to geeks.
Actually, there is a male equivalent to the body image distortion issues that women with eating disorders have, but instead of perceiving themselves as fat, men tend to perceive themselves as small (in more ways than one). So, a man is less likely to binge and purge, and more likely to overtrain at the gym and take steroids and various questionable herbal "male enhancement" concoctions.
In any case, playing a non-human character probably puts you out of the scope of what we're discussing, because no matter what you do, you cannot change your species. Playing a character that looks like you, but is a bit thinner is something that could possibly influence your image of your "true" self. Playing a troll, or a superintelligent shade of the colour blue isn't likely to affect your self-image in a simple way.
Wow, that was an incredible piece of irrelevant observation. Yes, indeed, if we moved the social security budget into space exploration, then the space program would be more interesting to geeks than it is now, because it would be developing more technology. You, however, seem to think I'm arguing geeks are interested in things that money is being spent upon. Obviously, that's not true, otherwise we'd be fascinated by federal flood insurance.
I'm just refuting the notion that geeks are necessarily militarists because they're interested in military technology. It's not the military aspect of military technology, it's the technology aspect, and to a lesser degree the controlled danger aspect. On the other hand, moving military technology money into Social Security, while it might be a good idea, doesn't make SS a geek obsession because there's no tech involved. Maybe if we someday go down the whole Japanese robots for elders path...
Well, I always tell young programmers that the things that matter most are at the top and at the bottom of the stack of the abstractions you are using. The in between part is always negotiable and subject to clever alterations.
Tetris is not apt to use a great deal of java's vast libraries, so if you're smart about loading just the parts you need, when you need them, why shouldn't Java on Javascript Tetris be fairly close to plain old Javascript Tetris, with a modest constant time hit. What I'd like to see is a non-trivial application that uses more than a few dozen Java classes load and run in a reasonable amount of time. You're either donwloading a significant portion of J2SE, encoded as javascript, or you're loading optimized javascript for some stack of method calls which works wonderfully for a small app but makes the problem worse for a large one.
I've seen some mobile development platforms that essentially analyze an app and create custom deployments of themselves for resource constrained devices. It's clever, but I'm not sure it works in the "drive by user" world of the web.
The only way the real costs are going to be hardware is if the software and content remains Free.
Well, that's an important point, but the way you posed it amounts to a catch-22. I think that the projects success demands free software and cheap hardware. I believe we agree on this.
We differ in this: you see the non-support of non-free software as a necessary condition for having free software. I don't agree. I am inferring that the project leadership thinks that supporting Windows is important to meeting the goals of cheap hardware. Clearly you disagree with the leadership if this is the case.
So which argument is more compelling? That non-free software support will kill the free software offerings, or that lack of the same will preclude amortizing the hardware costs across enough units to reach the required marginal costs? I can't prove one thing or the other without access to the planning numbers. I'm just proposing that possibly this is how Negroponte reads those numbers.
Of course he stuck his foot in it by using unpleasant names for people who might not as a group have been uniformly pleasant to him. That was very unwise of him.
Well, the military gets the money to buy the toys; if we moved, say, 20% of the military tech budget into space exploration, that would be a lot sexier right now.
This is one of the things the general public doesn't understand about nerds/geeks/whatever you call people who are defined by enthusiasms for difficult to understand things. Geeks differentiate between the utility of things as objects of study, and their intended utility, as any computer security researcher would tell you.
So just because you are fascinated by things that go boom doesn't mean you want to see them used on people. In fact, it's only idiots who like to play with explosives. The geek aspect of the game is doing things that would be stupid for other people, but not for you because you know exactly what is safe and what is not.
As far as geek militarism/pacifism goes, it seems to follow a pendulum like with everyone else, with the geeks being a bit ahead of the curve. You wouldn't be a geek exactly if your way of looking at things made you fit in.
Geeks are just smarter than the average populace. They aren't necessarily wiser. The very basis of wisdom is accepting that you might be wrong. Some people are so good at arguing and so used to being more right than their neighbor that they never have to confront their own fallibility. So geeks can represent both the best and worst humanity has to offer, the most enlightened viewpoints and the most stubbornly insular.
Perhaps not, but people whose needs are best met by Sugar are less likely to have their needs best met by Linux if the reason they can run Sugar on Windows is that resources that otherwise would have gone to making Sugar do what they needed on Linux were instead devoted to doing that on Windows.
This is the crucial argument, isn't it? But it's a pragmatic question. I don't think the usefulness of Linux to users is at issue, it's the usefulness of Sugar. If the project takes resources to make sure Sugar runs OK on XP, it doesn't threaten the utility of Linux, but the utility of Sugar.
It seems to me that the point fact that we're not discussing is that this machine clearly doesn't need XP support for any defensible technical reason. So it looks like putting any effort into getting Windows supported is a waste right?
Except the real costs of this program, if expanded to its envisioned potential, are going to be hardware. Hardware costs are at the center of this project, and everything depends on getting average hardware costs down by getting production volume up.
The reasons that the XP support is "needed" are clearly political. However the need for volume makes the need to overcome political challenges as real as the need to overcome battery limitations.
So, if you lose tens of thousands of units of volume because some education minister has been convinced that machines without Windows are useless, that's the real deal-breaking cost. So you give that country their Windows OLPC, and once they are successful they can see how little the Windows only part matters for themselves. And the problems that dealing with Windows licenses and support poses to expanding the program etc then become apparent.
The point is to get the problem out of the realm of the hypothetical into the realm practical decision making. Its a smart tactical move to give people what they want, then let them figure out the rest once it becomes a real decision for them.
Sure, but I'm assuming that you (a) want to arrive at the boat ramp in one piece and (b) you want to actually spend time with the boat in the water this weekend...
Well, sure, but I also remember when Japanese cars were called "rice cookers". I can recall marveling at this tiny Datsun sedan with its almost toylike 1.2l, 69 hp engine that one of the neighbors bought. I thought it was cool, although that was a decidedly non-cool idea of "cool".
The other neighbors had cool cars like the Plymouth Duster, which boasted 225hp and nearly five times the displacement of the Datsun. Another neighbor had a Buick Wildcat with a 401 inch (6.6l) V8 that generated an astonishing 325 hp, and that was without the optional supercharger. It's almost frightening to imagine that car, with its enormous mass and laughably primitive suspension supercharged. Some of the cars on my street had engines rated at significantly over 400 hp. Not a few had mpg ratings under 10 miles/gallon, and these weren't really considered big cars.
The point is, sometimes it pays to keep an eye peeled at the other extreme of the spectrum from everybody else when you're looking for sources of technological innovation. That doesn't necessarily mean you can pull your boat trailer with a 69 hp engine, of course.
Well, this is the old openness/freedom dichotomy, isn't it?
Arguably one of the most restrictive things about proprietary software is the way that the vendor's decisions are forced upon you. It's possible, for example, to use Microsoft's entire product stack and replace one item, say SQL server, or IIS. But really the path of least resistance is choose to be greater than 90% Microsoft centric or less than 10%.
Yes, it's important to have a completely open solution so that users can, if they wish, control their own destinies. But what happens when somebody decides that, for some reason, they need to have Windows XP on some of their machines? Do you use your control of the hardware and the Sugar ends of the sandwich to make it difficult for them to slip XP in between?
Of course not.
This whole thing reminds me of a speech I heard Barney Frank give on gay marriage, in which he ridiculed the notion that hordes of heterosexuals who had been happily married for years would suddenly decide they want to change sexual orientation because it would be so much nicer to marry somebody of the same sex as them.
Are people whose needs are best met by Linux going to abandon ship if they can run Sugar on Windows? Is Windows is so much more wonderful than Linux, that people secretly cherish a wish to run XP, but are stopped because they can't get the Linux only software they rely upon?
I think it's more likely that seeing the same hardware and applications running with different operating systems, users will be persuaded that the operating systems are things they can choose between.
I don't know a great deal about Sharepoint, other than as a user of IT department fielded Sharepoint facilities. They're OK, I guess, and they certainly look nice.
I wonder, though, about generalizing from the experience of supporting an office full of Notes email users. The thing about Notes is that the applications are scalable and manageable in a distributed fashion. Organizations were fielding distributed applications with tens of thousands of users on Notes back in the 1980s.
A lot of the arbitrary complexity of Notes administration has to do with the fact administration can be distributed. Of course, these days with Active Directory you can do that on MS tools as well, but I think the biggest problem with Notes was that it imposed too much of a learning curve on small deployments.
Slashdot Mirror
Actually, from the clips I have seen of the movie, Stark clearly has genuine human level (if not super-human) AI available.
He has a design support system consisting of virtual reality and robotics controlled by a sophisticated AI system. The AI system is capable of determining what its master intends to do, projecting its probable results, feeling concern, and expressing and explaining its disagreement.
That's beyond what a typical computer can do even in Star Trek.
I think the computational problem of translation is one that is amenable to the kinds of advances we've seen in the past few decades in computer chess. The big difference is that chess is a self-contained world, whereas language refers to things in the outside world that the speaker assumes the hearer knows in common with him. Even people do a pretty crappy job of understanding when people are coming from different backgrounds or even starting from different assumptions.
The experiential aspect of communication is probably going to be the greatest hurdle in high quality translation. "Real time" is must a matter of computational bandwidth.
Well, I'll debate the science merits of Harry Potter.
Magic, in Harry Potter, is depicted as much more scientific the science is in most sci-fi. It is more technological than mechanical engineering is in Iron Man.
It is true that magic, in Harry Potter, can only be done by some people. However if you are one of those people, it operates in a repeatable way. In fact, it is not just repeatable in a hocus pocus way, but in a way that allows people to abstract general laws and principles for its operation. Magical techniques and knowledge are improved both by practical experimentation and theoretical exploration. There are even scholarly journals in specialized magical sub-discplines ("Transfiguration Today", and "Challenges in Charming").
In fact in several cases these magical disciplines are actually real world cases of obsolete sciences that did not keep up with the empirical advances of science as a whole. Newton was an alchemist, after all, and numerology ("arithmancy") and astrology where related Hermetic proto-sciences much studied by Newton's contemporaries. Medicine barely squeaked over the line in the real world, leaving chiropractic just on the other side. Osteopathy was the last bit over the fence, and medicine is absorbing it into normal "allopathic" practice.
In Harry Potter's world economic competition spurs technological innovation in broomstick design. Research is discovering new treatments for ancient diseases (lycanthropy) and new applications for materials (dragon's blood). The government itself has advanced research facilities for advanced research into topics like the nature of time.
Education seems to follow this pattern: students are introduced to a theoretical concept, then they are given a practical demonstration, followed by a lab exercise. If they fail the lab exercise, they are given additional practice and theory assignments, culminating in an essay which demonstrates their grasp of the principles being taught. After a course of many such cycles, they are given an exam which determines their competency to study more advanced topics. In many ways, the educational practices of Hogwarts would be exemplary if it were an engineering college.
In most stories, magic is a kind of detritus of a past Golden Age; the knowledge of that age has been shattered and its pieces are slowly disappearing out of the world. Lucky (or sometimes very unlucky) individuals come into possession of bits of that knowledge, often in the form of ancient and hard-decipher scrolls or mouldy old grimoires. It is notable that Hogwarts texts, while at times a bit -- eccentric, are generally freshly printed and (sometimes) represent the efforts of reasonably contemporary scholars.
In short, the world of Harry Potter is most strongly characterized by progress in theoretical knowledge (science) and practical applications (technology). The one difference is that the minds of individual magic practitioners are interfaced in ways that are more direct than trough the the nerve endings in their bodies.
Well, not really. No way you can make a suit that flies that way for more than a few meters without anti-gravity. Where are the fuel tanks? And if a real person was inside of it, doing the things it supposedly does, you'd end up with a can of Campell's Hearty Tomato Stew with Real Meat Chunks.
And don't get me started on the issue of batteries. The existence of a battery system as good as Iron Man's is inconsistent with the world as we know it today.
What is different about the story is that it is located in a world that is more recognizably our own. The Batman movies are more blatantly archetypal. Gotham City is closer to Middle Earth than it is to our world. It isn't just that the story features real place names like "Vietnam" or "Afghanistan" (which in the story don't actually much resemble the real countries of the same name). It's that Tony Stark lives in a world where most people get along, and few like him do very well, by not rocking the boat, not questioning the status quo, and acting as if the aftermath of their actions were somebody else's job to clean up.
What gives the Iron Man story a different twist on most comic book stories is that if you strip away Tony Stark's engineering genius, he isn't really all that different from any other American; he just has the money to do indulge himself on a larger scale.
Well, consider this was 1964, it might have been a computed goto ...
In any case, this was not a language intended for software developers (like Algol). Nor was it a language intended for scientists (like Fortran). It was intended for CS students. Goto is desirable for demonstrating simple models of computation.
Dartmouth Basic had if/then, gosub, and for/next loops, which were much less gawdawful than Fortran's Do loops. It had fewer than a dozen functions, but they were well chosen to give students the ability to do interesting, non-trivial stuff.
I started programming maybe a dozen years after Dartmouth BASIC was introduced, and many of the interpreted language options available were not nearly so clean and well thought out.
Of course, that in part was the downfall of BASIC; it was well enough designed for its purpose that its impact was much larger than its target audience. Extending the language and libraries to such a wide variety of practical uses diluted the virtues of its original design. It was no longer a minimalistic language and runtime environment for students to learn and demonstrate concepts and for academics to do computations in. Its descendants became the patronizing choice when you targeted people who you assumed unable to learn anything better.
Or better yet ... Pakistan. After all Pakistan is an "ally" in the War on Terror.
I've used NTFS3G under fuse for mounting Windows partitions with virtual machines stored on them.
It's remarkably non-sucky, considering it is a user space reverse engineering of a proprietary FS. The maintainers haven't really even attempted to optimize the thing, and they were rewarded for their focus on correctness with -- not exactly outstanding performance, but surprisingly good performance under the circumstances.
The main drawback I can see is that it uses a lot of CPU for large file operations like copying multi-gb files.
Still, hats off to the developers for an exceptionally useful bit of software. I haven't found a single serious glitch in sharing NTFS partitions with Vista.
I've had good experiences with JFS and XFS.
I spent some time evaluating file systems for hosting virtual machine disks, and found JFS had a slight edge in XFS in consistency of performance, although XFS was perhaps a bit faster on average. To tell you the truth, even ext3 was quite usable. From a subjective standpoint the spread was not all that dramatic with the workloads I have.
In some ways, I remain very impressed with ext2/ext3, which was never designed to be the last thing in advanced file system. It was designed in 1993 to be a solid, all around performer and it remains so fifteen years later, which is amazing when you think about it. People say its performance is terrible, and that may well be true if you compare it to a file system with specialized strengths in the area of those strengths.
Still, I settled on JFS, which is probably the most rock-solid file system I've ever used. XFS has a reputation for losing data in power loss situations, which I haven't personally encountered. I have encountered data corruption on Reiser under similar conditions. For that reason I'd consider XFS or Reiser mainly if I thought I had a particular need their areas of exceptional competence.
I'm happy with jfs for all around use. It seems to be a very reasonable default choice, because it doesn't seem to have any glaring weaknesses. It performs solidly in a wide variety of situations, it handles heavy disk loads with low CPU usage.
Not a problem. Just let the user print bogus receipts.
At any point in the voting process, you can print a receipt. That receipt contains a cryptographic checksum of the voting, including the fact that the receipt was preliminary. Upon the actual casting of the vote a genuine receipt is printed.
Subsequently, if you want to verify a receipt, you need a court order to recover the hardware crypto key, which has been removed from the machine at the close of voting and placed at a secure escrow site.
True, but you don't have the scenario where the torturer is standing next to the prisoner with a cattle prod in his hand as the prisoner gets his email. The prisoner doesn't know whether it's actually his lawyer on the other end, and not an agent who has stolen the lawyer's key.
Prisoners in terror cases probably don't even have the right to send and receive unmonitored email. Prisoners don't have the right to make unmonitored calls.
Finally, you can't see that your client is haggard from sleep deprivation and rail thin from being starved and bruised from being beaten. The prisoner doesn't get the same reassurance that he hasn't disappeared into an anonymous void from exchanging emails.
I'm not saying that cryptography doesn't have some application in these situations -- although perhaps more for the coordination of the defense than for communicating with prisoners. I'm saying there is no substitute for being there.
Of course there was evidence. The evidence wouldn't have been enough to convict, but Reiser made tactical mistake of trying to bullshit his way out of it instead of leaving that to the closing arguments. All he had to do was leave the door open to reasonable doubt.
In the end, it is true that Reiser was convicted because he was arrogant. But while his arrogant demeanor probably didn't endear him to the jury, it was what his arrogance lead him to do that convicted him. His own testimony was evidence.
The system would be grossly unfair if you had to prove your innocence, but you don't. You don't have to have a theory that covers every thing. You just have to point out the holes in the prosecution's theories. Anybody who's watched reruns of Columbo knows that if you ad lib long enough, you get caught.
So he took the seat out of his car because he was sleeping in it. Didn't he notice there was an inch of water in the car? He didn't remember sleeping in a puddle of water. That's what got him convicted, he ran out of explanations and said something really, really stupid. What happens in the jury is that with twelve ordinary people, somebody is going to say, "well, maybe he planed to sleep in the car but didn't because it was full of water." And there you've got reasonable doubt. If you try to remove all doubt, you discredit the universe of possiblities under which you might be innocent.
In the end, Reiser's tragic flaw was that he was too arrogant to suppress his foolish impulses. Only an fool kills his wife to fix his problems. Even so, Reiser was smart enough to get away with it, but he was too arrogant to keep his mouth shut. If he had the kind of intelligence it took to restrain his mouth, he'd never have been in the situation where it needed restraining.
But this is demonstrably false as a general statement. Given that you must pay for both hardware and software, a software strategy which allows you to reduce your average cost spent on hardware could conceivably increase the total resources you have available, thus enabling you to spend more money supporting free software in the future.
I'm not saying it is necessarily so in this case, I couldn't without looking at the project's planning documents. I'm just saying that the way you budget for a product isn't quite like a household budget in which for the most part your discretionary costs are operational, not capital in nature.
Well, your argument would make sense if the plan was to abandon the idea of an entirely free software stack. Is that what Negroponte is advocating? I don't believe so.
Which is only possible to the degree that users depend on Windows only features. Since the shell is Sugar, the real question is whether users will become dependent on certain Windows only programs, and whether F/OSS replacements will not be available because of the monopoly effects of MS.
But I don't think that's very likely. Given the device's market, users will actually find the selection of Linux software much wider. This is just one of those things that customers think they absolutely have to have. After they have the things, find them useful, then they'll see their conception of the challenges was upside down. They'll ask, is there a chemistry instruction program for Windows? Yes, it cost $$$, but there's a free Linux one.
Sooner or later people will just ditch Windows and run Linux, because running Windows on this thing will obviously be stupid -- for the target audience. Heck, the kids can install Linux to get at software nobody will buy for them. They probably will. But they won't be able to if they don't have the hardware, and they won't have the hardware unless certain parties have their Windows security blanket.
Linux itself owes its existence to the desire to have something better to run on your cheap commodity hardware. Free software for the third world is dependent on there being hardware which means the conditions and means of third world people.
Not on route I95, you can't. In any case, a 150hp locomotive? What is that, a kiddy ride? I thought locomotives were in the 2000 - 6000 hp range, with self-propelled railcars being in the 500-1000 range. You must be measuring horsepower by some different method. I know that steam engines can be rated by boiler horsepower but be capable of delivering much higher net horsepower over short distances, by a factor of five or more. The old Stanley Steamers with a 20HP rating could actually deliver 125HP for a while, before boiler pressure started to drop.
Correction -- you might have been wrong. You might still be wrong now. Keep trying, and you'll get there.
Yes, but you missed the boat on why.
Actually, there is a male equivalent to the body image distortion issues that women with eating disorders have, but instead of perceiving themselves as fat, men tend to perceive themselves as small (in more ways than one). So, a man is less likely to binge and purge, and more likely to overtrain at the gym and take steroids and various questionable herbal "male enhancement" concoctions.
In any case, playing a non-human character probably puts you out of the scope of what we're discussing, because no matter what you do, you cannot change your species. Playing a character that looks like you, but is a bit thinner is something that could possibly influence your image of your "true" self. Playing a troll, or a superintelligent shade of the colour blue isn't likely to affect your self-image in a simple way.
Wow, that was an incredible piece of irrelevant observation. Yes, indeed, if we moved the social security budget into space exploration, then the space program would be more interesting to geeks than it is now, because it would be developing more technology. You, however, seem to think I'm arguing geeks are interested in things that money is being spent upon. Obviously, that's not true, otherwise we'd be fascinated by federal flood insurance.
I'm just refuting the notion that geeks are necessarily militarists because they're interested in military technology. It's not the military aspect of military technology, it's the technology aspect, and to a lesser degree the controlled danger aspect. On the other hand, moving military technology money into Social Security, while it might be a good idea, doesn't make SS a geek obsession because there's no tech involved. Maybe if we someday go down the whole Japanese robots for elders path...
Well, I always tell young programmers that the things that matter most are at the top and at the bottom of the stack of the abstractions you are using. The in between part is always negotiable and subject to clever alterations.
Tetris is not apt to use a great deal of java's vast libraries, so if you're smart about loading just the parts you need, when you need them, why shouldn't Java on Javascript Tetris be fairly close to plain old Javascript Tetris, with a modest constant time hit. What I'd like to see is a non-trivial application that uses more than a few dozen Java classes load and run in a reasonable amount of time. You're either donwloading a significant portion of J2SE, encoded as javascript, or you're loading optimized javascript for some stack of method calls which works wonderfully for a small app but makes the problem worse for a large one.
I've seen some mobile development platforms that essentially analyze an app and create custom deployments of themselves for resource constrained devices. It's clever, but I'm not sure it works in the "drive by user" world of the web.
Well, that's an important point, but the way you posed it amounts to a catch-22. I think that the projects success demands free software and cheap hardware. I believe we agree on this.
We differ in this: you see the non-support of non-free software as a necessary condition for having free software. I don't agree. I am inferring that the project leadership thinks that supporting Windows is important to meeting the goals of cheap hardware. Clearly you disagree with the leadership if this is the case.
So which argument is more compelling? That non-free software support will kill the free software offerings, or that lack of the same will preclude amortizing the hardware costs across enough units to reach the required marginal costs? I can't prove one thing or the other without access to the planning numbers. I'm just proposing that possibly this is how Negroponte reads those numbers.
Of course he stuck his foot in it by using unpleasant names for people who might not as a group have been uniformly pleasant to him. That was very unwise of him.
Well, the military gets the money to buy the toys; if we moved, say, 20% of the military tech budget into space exploration, that would be a lot sexier right now.
This is one of the things the general public doesn't understand about nerds/geeks/whatever you call people who are defined by enthusiasms for difficult to understand things. Geeks differentiate between the utility of things as objects of study, and their intended utility, as any computer security researcher would tell you.
So just because you are fascinated by things that go boom doesn't mean you want to see them used on people. In fact, it's only idiots who like to play with explosives. The geek aspect of the game is doing things that would be stupid for other people, but not for you because you know exactly what is safe and what is not.
As far as geek militarism/pacifism goes, it seems to follow a pendulum like with everyone else, with the geeks being a bit ahead of the curve. You wouldn't be a geek exactly if your way of looking at things made you fit in.
Geeks are just smarter than the average populace. They aren't necessarily wiser. The very basis of wisdom is accepting that you might be wrong. Some people are so good at arguing and so used to being more right than their neighbor that they never have to confront their own fallibility. So geeks can represent both the best and worst humanity has to offer, the most enlightened viewpoints and the most stubbornly insular.
This is the crucial argument, isn't it? But it's a pragmatic question. I don't think the usefulness of Linux to users is at issue, it's the usefulness of Sugar. If the project takes resources to make sure Sugar runs OK on XP, it doesn't threaten the utility of Linux, but the utility of Sugar.
It seems to me that the point fact that we're not discussing is that this machine clearly doesn't need XP support for any defensible technical reason. So it looks like putting any effort into getting Windows supported is a waste right?
Except the real costs of this program, if expanded to its envisioned potential, are going to be hardware. Hardware costs are at the center of this project, and everything depends on getting average hardware costs down by getting production volume up.
The reasons that the XP support is "needed" are clearly political. However the need for volume makes the need to overcome political challenges as real as the need to overcome battery limitations.
So, if you lose tens of thousands of units of volume because some education minister has been convinced that machines without Windows are useless, that's the real deal-breaking cost. So you give that country their Windows OLPC, and once they are successful they can see how little the Windows only part matters for themselves. And the problems that dealing with Windows licenses and support poses to expanding the program etc then become apparent.
The point is to get the problem out of the realm of the hypothetical into the realm practical decision making. Its a smart tactical move to give people what they want, then let them figure out the rest once it becomes a real decision for them.
Sure, but I'm assuming that you (a) want to arrive at the boat ramp in one piece and (b) you want to actually spend time with the boat in the water this weekend...
Well, sure, but I also remember when Japanese cars were called "rice cookers". I can recall marveling at this tiny Datsun sedan with its almost toylike 1.2l, 69 hp engine that one of the neighbors bought. I thought it was cool, although that was a decidedly non-cool idea of "cool".
The other neighbors had cool cars like the Plymouth Duster, which boasted 225hp and nearly five times the displacement of the Datsun. Another neighbor had a Buick Wildcat with a 401 inch (6.6l) V8 that generated an astonishing 325 hp, and that was without the optional supercharger. It's almost frightening to imagine that car, with its enormous mass and laughably primitive suspension supercharged. Some of the cars on my street had engines rated at significantly over 400 hp. Not a few had mpg ratings under 10 miles/gallon, and these weren't really considered big cars.
The point is, sometimes it pays to keep an eye peeled at the other extreme of the spectrum from everybody else when you're looking for sources of technological innovation. That doesn't necessarily mean you can pull your boat trailer with a 69 hp engine, of course.
Well, this is the old openness/freedom dichotomy, isn't it?
Arguably one of the most restrictive things about proprietary software is the way that the vendor's decisions are forced upon you. It's possible, for example, to use Microsoft's entire product stack and replace one item, say SQL server, or IIS. But really the path of least resistance is choose to be greater than 90% Microsoft centric or less than 10%.
Yes, it's important to have a completely open solution so that users can, if they wish, control their own destinies. But what happens when somebody decides that, for some reason, they need to have Windows XP on some of their machines? Do you use your control of the hardware and the Sugar ends of the sandwich to make it difficult for them to slip XP in between?
Of course not.
This whole thing reminds me of a speech I heard Barney Frank give on gay marriage, in which he ridiculed the notion that hordes of heterosexuals who had been happily married for years would suddenly decide they want to change sexual orientation because it would be so much nicer to marry somebody of the same sex as them.
Are people whose needs are best met by Linux going to abandon ship if they can run Sugar on Windows? Is Windows is so much more wonderful than Linux, that people secretly cherish a wish to run XP, but are stopped because they can't get the Linux only software they rely upon?
I think it's more likely that seeing the same hardware and applications running with different operating systems, users will be persuaded that the operating systems are things they can choose between.
think maybe you're supposed to notice these things could've been warheads rather than sattelites?
I'm just asking.
The Sharepoint point is an excellent one.
I don't know a great deal about Sharepoint, other than as a user of IT department fielded Sharepoint facilities. They're OK, I guess, and they certainly look nice.
I wonder, though, about generalizing from the experience of supporting an office full of Notes email users. The thing about Notes is that the applications are scalable and manageable in a distributed fashion. Organizations were fielding distributed applications with tens of thousands of users on Notes back in the 1980s.
A lot of the arbitrary complexity of Notes administration has to do with the fact administration can be distributed. Of course, these days with Active Directory you can do that on MS tools as well, but I think the biggest problem with Notes was that it imposed too much of a learning curve on small deployments.