You do realize that Motorola largest and splashiest phone release of the last year was the Moto MPX200, a Microsoft Smartphone-based phone? And the Nokia doesn't ship any Linux phones at all, and has no interest in doing so?
Well, actually, you don't need to hold all the files in ROM, only a very few. Those files need to be able to verify that the system is really running on the hardware it expects to be running on, and then make sure that the digital signatures on files which are loaded into any application on the computer are correct. That has the advantage that you can keep the number of files small, and the total amount of code necessary quite limited. That might allow you to never need to update the code: no UV flash involved, simply store it on the CPU itself.
That is the exactly the idea behind TCPA. If you're right, though, then Microsoft is also right, and has been right for quite a while. Yes, TCPA could be abused, but it also solves exactly the problem which it attempts to solve. The question is, is that a trade-off we all want to make?
In fact, the monoculture argument is used all the time against SMTP, just in different words. The difference is that the only way to fix a broken standard is to replace it. Microsoft argues that its operating systems are fixable. Whether or not that's true is still debatable, although the evidence support MS to date.
According to The Register, Munich is finding that trying to get Linux on the city's desktop is not yet possible -- even with direct help from IBM and SuSE. They're finding that what Microsoft has said about Word is true in general: it isn't just the big things that everybody uses which are a problem, but also the little things which a very small number of people can't do without. In that case of Word, it turns out that almost everybody has a few small, exotic features that they really need, and that those small features, taken together, add up to a much greater barrier than all the big features which everybody needs.
This isn't going to be the year of Linux on the desktop if that holds true generally.
Actually, the attack is more subtle than you think. The value of a random-words attack lies in the long-term damage it does to adaptive filters, not in how well or poorly it does with fixed filters.
When an adaptive filter sees a rare word in a spam, it is likely to assign that word high spamminess. Problem is, the next time you see that word is likely to be in a piece of ham, resulting in a false categorization of a piece of ham as spam. The user cost of such an assignment is very high, and so users will be forced to look at their junk mail...which is, after all, what the spammers want.
Oi. For a good popular intro, read Gleick's _Chaos_; it's not completely technically accurate, but it's pretty good. (And it paints the people involved really well.) Here's a very crude summary of the history -- I can come up with better references if people really want them.
Back in the early 60's, a meterologist named Ed Lorenz produced a model of part of the large-scale convection of Earth's atmosphere derived from the Navier-Stokes formulas for the flow of a viscous liquid in a heated pan. He wound up with wierd "numerical instability" in his simulations that led to some kind of unreproducibility. Now, Ed was the kind of bulldog of a scientist that wouldn't give up, and he finally concluded that what he was seeing was not due to his integrator, but, rather, implicit to the ODE itself. He published a paper in which he described "systematic sensistivity to initial conditions". Today, we call that "chaos in a dynamical system".
He was widely derided. To be fair to the larger community, he'd made a really big claim -- that the weather was unpredicatble outside of a finite interval -- and he didn't have proof of it; all he had were a bunch of simulation results of a subsystem. His data weren't enough to fully support the strength of his claim, and scientists are wisely exceptionally skeptical of such wild claims.
Ed was unusual, though, because it turns out that he was right anyway. Many years later, in the early eighties, as a result of the work of a bunch of people, it was shown that there did exist dynamical systems which were sensitive to initial conditions in exactly the way he'd described: an arbitrarily small perturbation would inevitably be exponentially magnified over time. One example is Ed's system, which contains the chaotic attractor he'd observed. That attractor is called the Lorenz Attractor in his honor.
The Lyapunov exponent of the attractor has been calculated, and it provides an upper bound on the period before which quantum mechanical noise due to Brownian motion in the atmosphere would become macroscopically detectible. That's the source of my 23 day figure.
In fact, von Neumann was provably wrong. Ed Lorentz' work on chaotic attractors in the Navier-Stokes system was so controversial presiely because it showed that long-term weather prediction over a period of more than about 23 days is impossible -- at least, if quantum mechanics is a valid theory.
The problem was that the MsFreePC site was an aggregator site, not a "facilitator" site. The submission was made via an electronic form to MsFreePC, which aggregated all the submissions it had collected and relayed them to Microsoft, using the vouchers to pay for the software Lindows had redistributed through MsFreePC. That was explicitly forbidden by the terms of the settlement.
It was a clever gambit. If Microsoft did choose to push back on Lindows, then Lindows would wind up with some very expensive publicity, and, if it was lucky, look like the underdog fighting the good fight. If MS failed to push back, then the company got a lot of money. It was a sleazy tactic, but clever.
As a matter of fact, yes, and that's one of the two things that this debate was all about. Electronic signatures were, in fact, not valid in this case (which isn't uncommon in settlements like this). Whether or not that is reasonable, though, there was a second problem. The settlement explicitly limited aggregators to collecting no more than a certain number of claims. MsFreePC was created on the basis of violating the terms of that settlement.
Most people preferentially free fuse cross-eyed: the right eye focuses on the left-hand image and vice versa. Some people, however, can free fuse in parallel: the right eye focuses on the right-hand image, the left eye on the left-hand image. Colleagues of mine who could do both told me that parallel fusion gave them less of a headache than cross fusion.
Mike Magee must be desperate for page hits: the author of the piece can't even count.
Microsoft only started breaking out the seven subunits about a year ago. During each of the quarters since then, three units -- not two -- have made money: client, Office, and server and tools. More than that, MSN (you know, the horrible money loser?) made money last quarter, and shows no signs of slowing revenue growth. That's four of seven making money, not two.
The author of the Inquirer piece would like to lump the two OS divisions together, but that makes no sense: F/OSS systems don't compete against the client yet, only against the server and tools segment. Revenue in that segment is growing faster than the segment. That's not being beaten by Linux; it competing solidly, despite a price disadvantage.
Worse, for the author's thesis, the handhelds division is hardly "losing money fast" -- instead, it's losing money at a constant rate, with its revenues more than doubling each year. If current patterns continue, that division will be profitable in the current quarter or the next quarter. That's not clearly going to happen, but it certainly doesn't seem unlikely.
That leaves two divisions not making money: Home and Small business solutions. Those are both new businesses for Microsoft, and they're both businesses where Microsoft expects to lose money for about a decade, just as it did with servers, with MSN, and with handhelds.
But, hey, the story predicts the death of the internet...I mean, the death of Microsoft. SO we've got to front page it to give Magee and/. extra ad hits.
Hiring a code maintainer is exatly what a company does not want to do. Maintainers are very expensive, and you can't hire them in pieces. The closest you can come to this is to hiring a company which specializes in maintaining the package you wanted. In that case, you've lost the price advantage you had, and you've not gained the source. You don't have people on site who can actually read the source, and if your service provider goes under, you're left without an escape.
What's the advantage of this over buying from Microsoft?
Actually, Florida has both alligators and crocodiles. The Florida Crocodile is an endangered species. Almost all of the known specimens are currently found in the downstream heat plume of one of the local nuclear power plants.
(No, I'm not joking. Nobody knows why the crocs have congregated there, but it seems to work for them. Works for the humans, too: alligators don't tend to bother people if they're not bothered first. Crocodiles are considerably more aggressive, so having them stay in a place where people can be easily convinced to avoid is a good thing.)
If you've ever been in a long multi-person thread, you know that writers will sometimes respond to more than one message in a single response. More than that, they're change the subject when the subject of their particular message is different from the rest of the conversation. This makes their e-mails more effective at communicating with the other people involved.
More than that, this research has applications to recognizing the relationship among different mails in my inbox without being limited to the things which a computer can recognize with simple pattern matching. That's useful: searching my mail store is a huge chore unless I know exactly what I'm looking for. Unfortunately, I need to search precisely when I only remember the general outlines of a conversation.
Not really. Outlook and other mail clients look at the mail header to define a conversation. This is intended to do a better job than that: it looks at the content of the mail as well and tries to infer the threading structure from that.
Since, in this case, the patch actually introduces an exploitable buffer overrun, I'd suggest that your standards of trust are somewhat too low. (Not to mention that it makes an unnecessary access to these clown's web site, makes invalid assumptions about URL structure, and also leaks memory like a stuck pig, so that the copy of explorer on your system will degrade rapidly over time.)
Do yourself a favor: don't install this, and don't encourage anybody else to do so, either.
Welll....not exactly. If the craft were to hit the surface of any Europa (wildly unlikely) and if Europa actually has liquid water oceans with ice-covered surfaces (not proven), then the reactor would melt through the ice, boil a large volume of water, and then sink to the bottom of the ocean, which would contaminate the deep structures of the moon more than the Jovian wind does.
However, (a) a reactor is a total nit on the scale of Europa, so the damage would be extremely localized, and (b) the moon itself is sufficiently tectonically active, due to tidal forces, that the reactor would be quickly swallowed up by the exolounar core, thus reducing its effects even more.
Bottom line: it'd be a catastrophe, but not one as large as it appears at first.
Your argument is a lot stronger when it comes to biological contamination, though. I haven't pushed the numbers, and I think that even a couple of hours in the Jovian magnetosphere ought to be sufficient to kill any unshielded terrestrial life forms which had contaminated the probe during assembly. I certainly hope so.
Slashdot Mirror
can she read it?
You do realize that Motorola largest and splashiest phone release of the last year was the Moto MPX200, a Microsoft Smartphone-based phone? And the Nokia doesn't ship any Linux phones at all, and has no interest in doing so?
Well, actually, you don't need to hold all the files in ROM, only a very few. Those files need to be able to verify that the system is really running on the hardware it expects to be running on, and then make sure that the digital signatures on files which are loaded into any application on the computer are correct. That has the advantage that you can keep the number of files small, and the total amount of code necessary quite limited. That might allow you to never need to update the code: no UV flash involved, simply store it on the CPU itself.
That is the exactly the idea behind TCPA. If you're right, though, then Microsoft is also right, and has been right for quite a while. Yes, TCPA could be abused, but it also solves exactly the problem which it attempts to solve. The question is, is that a trade-off we all want to make?
In fact, the monoculture argument is used all the time against SMTP, just in different words. The difference is that the only way to fix a broken standard is to replace it. Microsoft argues that its operating systems are fixable. Whether or not that's true is still debatable, although the evidence support MS to date.
According to The Register, Munich is finding that trying to get Linux on the city's desktop is not yet possible -- even with direct help from IBM and SuSE. They're finding that what Microsoft has said about Word is true in general: it isn't just the big things that everybody uses which are a problem, but also the little things which a very small number of people can't do without. In that case of Word, it turns out that almost everybody has a few small, exotic features that they really need, and that those small features, taken together, add up to a much greater barrier than all the big features which everybody needs.
This isn't going to be the year of Linux on the desktop if that holds true generally.
Oh, cool! When you did the fake Nigerian spam, did you use the bigram method as a guide?
Actually, the attack is more subtle than you think. The value of a random-words attack lies in the long-term damage it does to adaptive filters, not in how well or poorly it does with fixed filters.
When an adaptive filter sees a rare word in a spam, it is likely to assign that word high spamminess. Problem is, the next time you see that word is likely to be in a piece of ham, resulting in a false categorization of a piece of ham as spam. The user cost of such an assignment is very high, and so users will be forced to look at their junk mail...which is, after all, what the spammers want.
Oi. For a good popular intro, read Gleick's _Chaos_; it's not completely technically accurate, but it's pretty good. (And it paints the people involved really well.) Here's a very crude summary of the history -- I can come up with better references if people really want them.
Back in the early 60's, a meterologist named Ed Lorenz produced a model of part of the large-scale convection of Earth's atmosphere derived from the Navier-Stokes formulas for the flow of a viscous liquid in a heated pan. He wound up with wierd "numerical instability" in his simulations that led to some kind of unreproducibility. Now, Ed was the kind of bulldog of a scientist that wouldn't give up, and he finally concluded that what he was seeing was not due to his integrator, but, rather, implicit to the ODE itself. He published a paper in which he described "systematic sensistivity to initial conditions". Today, we call that "chaos in a dynamical system".
He was widely derided. To be fair to the larger community, he'd made a really big claim -- that the weather was unpredicatble outside of a finite interval -- and he didn't have proof of it; all he had were a bunch of simulation results of a subsystem. His data weren't enough to fully support the strength of his claim, and scientists are wisely exceptionally skeptical of such wild claims.
Ed was unusual, though, because it turns out that he was right anyway. Many years later, in the early eighties, as a result of the work of a bunch of people, it was shown that there did exist dynamical systems which were sensitive to initial conditions in exactly the way he'd described: an arbitrarily small perturbation would inevitably be exponentially magnified over time. One example is Ed's system, which contains the chaotic attractor he'd observed. That attractor is called the Lorenz Attractor in his honor.
The Lyapunov exponent of the attractor has been calculated, and it provides an upper bound on the period before which quantum mechanical noise due to Brownian motion in the atmosphere would become macroscopically detectible. That's the source of my 23 day figure.
In fact, von Neumann was provably wrong. Ed Lorentz' work on chaotic attractors in the Navier-Stokes system was so controversial presiely because it showed that long-term weather prediction over a period of more than about 23 days is impossible -- at least, if quantum mechanics is a valid theory.
No, there was no limit to the number of claims.
The problem was that the MsFreePC site was an aggregator site, not a "facilitator" site. The submission was made via an electronic form to MsFreePC, which aggregated all the submissions it had collected and relayed them to Microsoft, using the vouchers to pay for the software Lindows had redistributed through MsFreePC. That was explicitly forbidden by the terms of the settlement.
It was a clever gambit. If Microsoft did choose to push back on Lindows, then Lindows would wind up with some very expensive publicity, and, if it was lucky, look like the underdog fighting the good fight. If MS failed to push back, then the company got a lot of money. It was a sleazy tactic, but clever.
As a matter of fact, yes, and that's one of the two things that this debate was all about. Electronic signatures were, in fact, not valid in this case (which isn't uncommon in settlements like this). Whether or not that is reasonable, though, there was a second problem. The settlement explicitly limited aggregators to collecting no more than a certain number of claims. MsFreePC was created on the basis of violating the terms of that settlement.
From Yahoo Finance, in regard to SCOX:
% Held by Insiders: 45.83%
% Held by Institutions: 31.94%
That looks like the exact opposite of a pump and dump style investment strategy.
Most people preferentially free fuse cross-eyed: the right eye focuses on the left-hand image and vice versa. Some people, however, can free fuse in parallel: the right eye focuses on the right-hand image, the left eye on the left-hand image. Colleagues of mine who could do both told me that parallel fusion gave them less of a headache than cross fusion.
Oh, wait -- what would I expect him to say about an OS which uses a demon as its logo. Sorry.
A GE bathtub toaster? I don't know if that's the worst invention of 2003, but it's certainly one of the strangest. Why would I want a toasted bathtub?
You do know that GRE tunnels are the generic name for PPTP, don't you?
Mike Magee must be desperate for page hits: the author of the piece can't even count.
/. extra ad hits.
Microsoft only started breaking out the seven subunits about a year ago. During each of the quarters since then, three units -- not two -- have made money: client, Office, and server and tools. More than that, MSN (you know, the horrible money loser?) made money last quarter, and shows no signs of slowing revenue growth. That's four of seven making money, not two.
The author of the Inquirer piece would like to lump the two OS divisions together, but that makes no sense: F/OSS systems don't compete against the client yet, only against the server and tools segment. Revenue in that segment is growing faster than the segment. That's not being beaten by Linux; it competing solidly, despite a price disadvantage.
Worse, for the author's thesis, the handhelds division is hardly "losing money fast" -- instead, it's losing money at a constant rate, with its revenues more than doubling each year. If current patterns continue, that division will be profitable in the current quarter or the next quarter. That's not clearly going to happen, but it certainly doesn't seem unlikely.
That leaves two divisions not making money: Home and Small business solutions. Those are both new businesses for Microsoft, and they're both businesses where Microsoft expects to lose money for about a decade, just as it did with servers, with MSN, and with handhelds.
But, hey, the story predicts the death of the internet...I mean, the death of Microsoft. SO we've got to front page it to give Magee and
Hiring a code maintainer is exatly what a company does not want to do. Maintainers are very expensive, and you can't hire them in pieces. The closest you can come to this is to hiring a company which specializes in maintaining the package you wanted. In that case, you've lost the price advantage you had, and you've not gained the source. You don't have people on site who can actually read the source, and if your service provider goes under, you're left without an escape.
What's the advantage of this over buying from Microsoft?
Nah -- you need laser beams for their heads, too. The reactor's the easy part; fitting the d*ned lasers on is a real chore.
You'd probably do better to stick with frikken sharks.
Yeah, and you had to walk three miles uphill each way in the snow to learn it in school.
Actually, Florida has both alligators and crocodiles. The Florida Crocodile is an endangered species. Almost all of the known specimens are currently found in the downstream heat plume of one of the local nuclear power plants.
(No, I'm not joking. Nobody knows why the crocs have congregated there, but it seems to work for them. Works for the humans, too: alligators don't tend to bother people if they're not bothered first. Crocodiles are considerably more aggressive, so having them stay in a place where people can be easily convinced to avoid is a good thing.)
No, it isn't just silly.
If you've ever been in a long multi-person thread, you know that writers will sometimes respond to more than one message in a single response. More than that, they're change the subject when the subject of their particular message is different from the rest of the conversation. This makes their e-mails more effective at communicating with the other people involved.
More than that, this research has applications to recognizing the relationship among different mails in my inbox without being limited to the things which a computer can recognize with simple pattern matching. That's useful: searching my mail store is a huge chore unless I know exactly what I'm looking for. Unfortunately, I need to search precisely when I only remember the general outlines of a conversation.
Not really. Outlook and other mail clients look at the mail header to define a conversation. This is intended to do a better job than that: it looks at the content of the mail as well and tries to infer the threading structure from that.
Since, in this case, the patch actually introduces an exploitable buffer overrun, I'd suggest that your standards of trust are somewhat too low. (Not to mention that it makes an unnecessary access to these clown's web site, makes invalid assumptions about URL structure, and also leaks memory like a stuck pig, so that the copy of explorer on your system will degrade rapidly over time.)
Do yourself a favor: don't install this, and don't encourage anybody else to do so, either.
Welll....not exactly. If the craft were to hit the surface of any Europa (wildly unlikely) and if Europa actually has liquid water oceans with ice-covered surfaces (not proven), then the reactor would melt through the ice, boil a large volume of water, and then sink to the bottom of the ocean, which would contaminate the deep structures of the moon more than the Jovian wind does.
However, (a) a reactor is a total nit on the scale of Europa, so the damage would be extremely localized, and (b) the moon itself is sufficiently tectonically active, due to tidal forces, that the reactor would be quickly swallowed up by the exolounar core, thus reducing its effects even more.
Bottom line: it'd be a catastrophe, but not one as large as it appears at first.
Your argument is a lot stronger when it comes to biological contamination, though. I haven't pushed the numbers, and I think that even a couple of hours in the Jovian magnetosphere ought to be sufficient to kill any unshielded terrestrial life forms which had contaminated the probe during assembly. I certainly hope so.