You must be from Microsoft, and this simple truth of people's experiences with Vista hurts. Well tough. Vista is bug-ridden like Windows was until XP, and by abandoning XP for a new O/S, MS has several years of bug-fixing ahead of it before Vista reaches XP standards. Instead of wasting time trying to dismiss people's troubles with Vista, why don't you do something more productive, like fixing the code? "You say the gap between the rich and poor is unacceptable, but communists also say this; therefore you are a communist!"
Oh, dear. Have a flip through the Wikipedia article on Association fallacy. It is logical thinking of a rather strange and fuzzy kind that believes that if a news outlet is criticised for neglecting to uphold even the most basic standards on reliable sources, on a story that happens to be anti-Microsoft, it follows that the criticiser must be an employee of Microsoft (I'm not).
Although I must congratulate you on the way you started that paragraph ("You must be from Microsoft, and this simple truth of people's experiences with Vista hurts. Well tough."): one of the most expediently erected straw man I have since for months.
Well one would be to design an OS that needs fewer reboots to fix problems.
Unless I get a kernel update my Linux box doesn't need to reboot.
Most drivers should be designed to be updated without a reboot. Some like the mass storage my need to reboot.
With a properly designed modular operating system you should rarely need to reboot.
It goes back to FIX WINDOWS! ...Or not. A quick check of the Linux kernel updates list reveals that the most recent stable version was declared on the 15th August; the one before that, 9th August; before that, 4th August. Even if you don't update every single one (which, to be fair, most people probably don't), it doesn't exactly cast MS's once-a-month reboots in a particularly poor light by comparison.
But in any case, the frequency is irrelevent. As you say, "Unless I get a kernel update my Linux box doesn't need to reboot" -- but when you do get a kernel update, you do need to reboot, and so does everyone else running Linux, especially if the update patches some critical security hole that might otherwise be exploited. And so the situation if everyone was running Linux remains unchanged from what it is now: there are always going to be some situations where critical kernel security updates are released, and there is always going to be mass rebooting at that time, and any TFA-like bug for which that causes a problem will have a problem. "An OS that needs fewer reboots to fix [the] problem" would actually not fix the problem; it would merely delay the problem -- so maybe the Skype network would not have gone down until the next critical kernel security update, but it would still have eventually.
The fact is there is no desktop-level "solution" because there is no desktop-level problem. Any network like Skype's is always going to have to handle mass network repropogation eventually; the fact that it couldn't was a bug in Skype's algorithm, as TFA says.
If the only source for a story is one which would not qualify as a Reliable Source per Wikipedia guidelines, reconsider whether you have a story.
Yes, that includes stories based on a thread on a hardware forum.
I'm not asking for Slashdot to be held to journalistic standards (multiple source and/or independant investigative reporting). But Slashdot is supposed to be a news site, not a rumour mill. Is a single reliable source for unverified speculation like this too much to ask?
There isn't that is what so terrible is that once a month or so a huge percentage of systems are being forced to reboot! I do believe I already asked: what is the alternative?
You seem to be implying that Patch Tuesday is the problem, but if anything, patch Tuesday helps alleviate the problem: it means that the mass reboot cycle only happens once a month, rather than whenever a reboot-requiring-update comes out, which could be several times per month.
To reiterate what I said before: If a patch that requires rebooting is released, all Automatic-Update-enabled Windows boxes are going to need to reboot at some point. The only alternative to them all doing it on the same day is to stagger the release of the patch over a number of days, which is a very bad idea indeed: malicious crackers could analyse the patch on the first of those days, release an exploit for whatever the patch fixes immediately, and be guaranteed a few days of millions of unpatched systems to wreak havoc.
ven then if you think about it the idea that EVERY windows system is going to have to reboot on a certian day is just laughable. But then what's the alternative? If a patch that requires rebooting is released (whether on the day it's written or patch Tuesday is irrelevent), all AU-enabled Windows boxes are going to need to reboot at some point. The only alternative to them all doing it on the same day is to stagger the release of the patch over a number of days, which is a very bad idea indeed: malicious crackers could analyse the patch on the first of those days, release an exploit for whatever the patch fixes immediately, and be guaranteed a few days of millions of unpatched systems to wreak havoc.
...Huh? Why would this affect MS in the least, let alone strike "a huge blow against Windows on the wordstation"? It's not as if this will happen every month: It was a one off due to a bug in Skype's network algorithms, it's already been fixed, and the chances of it happening again are negligibe.
Arent people usually complaining that windows userd doesnt install the security patches? Note since XP SP2, which turned automatic updates on by default.
now people complain that they actually DO install them... WHEN OH WHEN is people satified? Perhaps the people complaining that AU is on by default are different people to the ones who were complaining when it wasn't? Just possibly the former set were the ones who argued that AU shouldn't be on by default, and the latter set, the ones who argued that it should? Just a thought.
I can't remember where it was, but I remember reading a few years ago (before SP2, and thus before automatic updates turned on by default) a very convincing argument which statistically analysed the spread of some virus, which exploited a vulnerability that had been patched in Windows Update quite a few months before, to argue that a majority of home users never update their system. Automatic updates switched on by default was probably the single most important thing SP2 did, even more than turning on the firewall by default. (Of course, even now, a fairly big proportion still don't have AU on, since they never installed SP2 because they never update their system... )
be fair to Skype you have to admit that 85% of the world's computers turning off at the same time is not an event a normal person would predict "Not an event a normal person could predict"? Ummm, not even when that same event has happened on the second Tuesday of every month for nearly a decade?
RTFA: the only reason that this month was different was a bug in Skype's network allocation algorithm, a bug of a type that was triggered by a normal Patch Tuesday mass rebooting event.
Maybe start making my own hardware like Apple to cut down on headaches and customer support. They've got experience with the Xbox. Maybe making their own computers isn't such a bad idea. Guarantee the OS will work on THEIR machine, and if someone wants to try it on something cheaper from Dell, go nuts, but you're on your own. It would certainly solve a hell of a lot of problems for MS, but there's a showstopper: for a convicted software monopolist like Microsoft, making your own desktop PC hardware would be Anticompetitive Practices on a scale way, way, way bigger than bundling IE or WMP ever was. The EU and DoJ would be down like a ton of bricks. If Microsoft wasn't split up before by the DoJ because of the current US administration's support for big business, as some on this forum believe; with practically every OEM on the planet (which includes most of the big-name electronics companies) clamoring for Microsoft's head that situation would about-turn faster than you can say "antitrust".
Even if the EU & DoJ didn't exist, Microsoft is what it is today in a large part through the use of backscratch ("I'll scratch your back if you scratch mine") exclusivity agreements with all the major OEMs. MS introducing its own hardware would be regarded by the OEMs as a severe breach of "trust" (for want of a better word), and MS does not want the CEO of Sony to start wondering whether the time of Linux on the Desktop has come...
Which is all actually rather a pity, since Microsoft hardware (mice, keyboards, XBoxes etc.) is actually pretty good, and the retail PC market could well to with a bit of operating system diversification; but such is life...
How much different is what's in an Apple box compared to... say... a Dell? Not much. But Apple *knows* what's in an Apple box. Microsoft? There are a good 100 different motherboard manufacturers (most of them 'no-name-brand' Chinese/Taiwanese/S.Korean manufacturers), each of which has on average at least 10 different models. Ditto memory. Ditto graphics cards. Etc., etc. The only major exception is CPU companies, due to the sheer investment it takes to make something that complex. The number of different combinations is ridiculous, and clearly untestable, even given all the resources in the world. And for the most part, each manufacturer writes their own drivers. And when you think how bad drivers from big-name brands like Creative and ATi are, a lot of no-brand drivers are worse. E.g. crap like this.
Whilst the situation is a bit better with major OEMs (e.g. you don't tend to get stuff like this, it's not much. The Vista Certified etc. stickers are not saying that that *particular* hardware configuration has been tested by MS (or, indeed, anyone), all they're saying is that the specification meets the recommended MS guidelines for Vista. Because -- here's a key point -- the actual hardware configuration will change regularly. A Sony ordered on Tuesday might have a no-name-brand A motherboard, whilst the identical machine ordered on Thursday might have a no-name-brand B one. The same spec, of course, since with the big-name-brands the actual motherboard manufacturer is likely making that entire line of motherboards specifically for Sony, to Sony's specifications; but the actual manufacturer and hardware is definitely not fixed. An analogy might be KFC: the product you see is pretty homogenised, but the chickens do not all come from one vast, centralised farm; they're bought from many hundreds of seperate farms, and the farm that the chicken you're eating will have come from will vary from week to week.
All this so far is generalised Microsoft/Linux model vs Apple model, rather than specific to Vista problems, but frankly most people here seem to have short memories: XP experienced vast amounts of hardware compatibility problems when it was first introduced, as did 2000, NT (especially), 98, 95, 3.1, etc. (I don't knw whether anyone's been able to tell whether ME did, since it had so many problems: who could tell which were due to hardware compatibility..?). Largely, this will get better with time (over a timescale of years, not months) -- part of the problem, I think, is that a lot of manufacturers are just using the same identical drivers they designed for XP and expecting them to work; and they often do, but sometimes don't (as happened with 2000 to XP, 95 to 98, etc.).
The more efficient one would have "no waste" heat to dispose of... Google "Second law of Thermodynamics". Also "Heat engine" and "Carnot cycle". There will always be waste heat.
I'm afraid Wonko's right. The total efficiency converges to Carnot's. If there is any *usable* waste heat left at the end of a cycle to put into another heat engine, then the first cycle wasn't running at full (reversible) Carnot efficiency.
BTW, if a heat engine were ever to be used in a closed system, then its efficiency would quickly converge to 0%, since the hot source would cool down and the cold source would heat up! The Carnot cycle, as I said in my other post, assumes infinite, constant temperature hot and cold sources, i.e. effectively the same as a system where heat is constantly added to the hot source (by a nuclear reactor) and taken away from the cold source (by a running river) to maintain a constant temperature.
The cold sink is river's water. Even if it is at the boiling point, T_c is under 300. T_h is the temperature of the nuclear reaction, which could be in the thousands. It is not there because of the engineering limits. The T_c/T_h ratio can thus be well under 10% allowing for over 90% efficiency. 0C is just over 273K, so for water at boiling point T_c would be 373, but that's a minor point. The link you gave actually specified, for that type of reactor, a T_h of 1000C, not "in the thousands"; which, even assuming T_c is at atmospheric temperature (20C), would have a maximum theoretical efficiency of 77%, rather than 90%. Moreover, remember that that assumes an infinitesimal carnot cycle (one with a negligible temperature gradient, i.e., assumes all processes are reversible), which is clearly not very realistic: as soon as you drop that assumption, you get the more realistic efficiency of only 52% (from 1 - sqrt(T_c/T_h) ).
That formula (of the Carnot cycle's efficiency) applies to closed systems, which a power plant is not. Because the "supply of cold" is external and endless (a river), the limit does not apply. No. The equation used (for the maximum theoretical efficiency of a Carnot cycle) already assumes infinitely large hot and cold sources, at constant temperatures. A river, since it is as you say eternal and endless, is actually very well modelled by this.
For example, the high-temperature steam/water ("waste heat") can be used as the heat source for another heat engine (of lesser efficiency)... Again, not really. At the maximum theoretical efficiency (reversible Carnot cycle), there is no usable waste heat: the output into the cold sink at any point is negligibly higher than the temperature of the sink. The only reason there is usable waste heat in current power stations is that they don't operate as reversible Carnot cycles (1 - T_c/T_h), they operate, broadly, as endoreversible cycles, which have a theoretical efficiency of, as I said above, 1 - sqrt(T_c/T_h). Adding a chain of progressively less efficient heat engines to make use of the waste heat might well bring the efficiency from 1 - sqrt(T_c/T_h) to near 1 - T_c/T_h, but it will never exceed 1 - T_c/T_h; that is a fundamental limit.
No, dear, I did not. You are struggling with the Second Law of Thermodynamics [...] the only "fundamental limit of thermodynamics" is that the ratio be below 1... Yes, dear, I'm afraid you did. It is you who are still struggling with the Second Law, and the concept of a heat engine. The "fundamental limit of thermodynamics" for a heat engine is not 1, but rather 1 - T_c/T_h, where T_c and T_h are the absolute temperatures of the cold and hot sinks, respectively. A heat engine running at this efficiency would give a net entropic change of zero: any more efficient, and the second law of thermodynamics would be violated.
Sorry about that. Post should, of course, have read:
Ummm... Seriously? Of the seven toolbars in Word 2007, one is labelled "Mailings". Sure enough, mail merge is the third icon on that toolbar. If you seriously think that having mail merge under "Mailings" is less intuitive than having under "Tools" (aka "miscellaneous"), then, well... You're entitled to your opinion...
At the risk of getting trounced on by all the scientists here, i disagree with Einstein's theory about the speed of light... If you take Einstein's two postulates as axioms, the rest of Special Relativity indisputibly mathematically follows. Thus, if Special Relativity is wrong, one of the two postulates must be incorrect.
In case you don't know, the two postulates are:
First postulate - The laws of physics are the same in all inertial frames of reference. In other words, there are no privileged inertial frames of reference.
Second postulate - The speed of light in a vacuum is a constant, which is independent of the motion of the light source.
So, which one of them do you think is wrong? If you're right, one of them must be...
Ok. So I'm still not convinced that a photon has no mass. It exists, has volume and is detecable as a particle/atom or whatever. Let's take those one by one.
- It exists. Yup.
- It has volume. Nope. It has one measureable spatial quality, which is wavelength. To ask what the other two dimensions of a photon are ('volume' implies three) is meaningless. Remember, a photon is a packet of energy, a propogating electromagnetic wave (the electric and magnetic components at right angls to each other). It has no well-defined volume.
- It is "detectable as a particle/atom or whatever". Ummm... Oh dear. A particle, in some sense, certainly: google "wave-particle duality". An atom, no.
To get back to the issue at hand, the thing which might be confusing you is that is has no rest-mass. If it *did* have rest-mass, it couldn't move at the speed of light. But it certainly does have mass in the sense that it is moving, so has kinetic energy, and, as Einstein pointed out, energy and mass are just two words for the same thing. It's energy due to its momentum is given by E=h*frequency=hc/wavelength.
But I *do* know that you collapsed yours. No, you don't. How could you possible know? All you know is that you have a particle, and you can measure its spin in either of two axis. Now, say you measure it in the left-right axis and its spin comes up as left. What do you know now? You do know that if the corresponding entangled particle has been measured in the left-right axis, it would have come up as right. But this does not tell you whether it has actually been measured. There is no way to tell whether the other party has measured their particle. No information has been transferred.
Caveat: you could, of course, arrange to, say, have the other party shoot their pet cat if their result comes up 'left', and repreive it if 'right'. By measuring your particle, you could then know whether the cat is dead or alive. But in this case, no information is actually transferred across the entanglement: the situation is the same as if you had carried a sealed envelope with you written by the other party, with whether they are planning to kill their cat written inside it, that you then open at a later time. You know whether the cat is dead or not, but again, information is not of course instantaneously transferred from the other party to you.
Slashdot Mirror
Oh, dear. Have a flip through the Wikipedia article on Association fallacy. It is logical thinking of a rather strange and fuzzy kind that believes that if a news outlet is criticised for neglecting to uphold even the most basic standards on reliable sources, on a story that happens to be anti-Microsoft, it follows that the criticiser must be an employee of Microsoft (I'm not).
Although I must congratulate you on the way you started that paragraph ("You must be from Microsoft, and this simple truth of people's experiences with Vista hurts. Well tough."): one of the most expediently erected straw man I have since for months.
But in any case, the frequency is irrelevent. As you say, "Unless I get a kernel update my Linux box doesn't need to reboot" -- but when you do get a kernel update, you do need to reboot, and so does everyone else running Linux, especially if the update patches some critical security hole that might otherwise be exploited. And so the situation if everyone was running Linux remains unchanged from what it is now: there are always going to be some situations where critical kernel security updates are released, and there is always going to be mass rebooting at that time, and any TFA-like bug for which that causes a problem will have a problem. "An OS that needs fewer reboots to fix [the] problem" would actually not fix the problem; it would merely delay the problem -- so maybe the Skype network would not have gone down until the next critical kernel security update, but it would still have eventually.
The fact is there is no desktop-level "solution" because there is no desktop-level problem. Any network like Skype's is always going to have to handle mass network repropogation eventually; the fact that it couldn't was a bug in Skype's algorithm, as TFA says.
Can I float an idea past the Slashdot editors:
If the only source for a story is one which would not qualify as a Reliable Source per Wikipedia guidelines, reconsider whether you have a story.
Yes, that includes stories based on a thread on a hardware forum.
I'm not asking for Slashdot to be held to journalistic standards (multiple source and/or independant investigative reporting). But Slashdot is supposed to be a news site, not a rumour mill. Is a single reliable source for unverified speculation like this too much to ask?
You seem to be implying that Patch Tuesday is the problem, but if anything, patch Tuesday helps alleviate the problem: it means that the mass reboot cycle only happens once a month, rather than whenever a reboot-requiring-update comes out, which could be several times per month.
To reiterate what I said before: If a patch that requires rebooting is released, all Automatic-Update-enabled Windows boxes are going to need to reboot at some point. The only alternative to them all doing it on the same day is to stagger the release of the patch over a number of days, which is a very bad idea indeed: malicious crackers could analyse the patch on the first of those days, release an exploit for whatever the patch fixes immediately, and be guaranteed a few days of millions of unpatched systems to wreak havoc.
...Huh? Why would this affect MS in the least, let alone strike "a huge blow against Windows on the wordstation"? It's not as if this will happen every month: It was a one off due to a bug in Skype's network algorithms, it's already been fixed, and the chances of it happening again are negligibe.
I can't remember where it was, but I remember reading a few years ago (before SP2, and thus before automatic updates turned on by default) a very convincing argument which statistically analysed the spread of some virus, which exploited a vulnerability that had been patched in Windows Update quite a few months before, to argue that a majority of home users never update their system. Automatic updates switched on by default was probably the single most important thing SP2 did, even more than turning on the firewall by default. (Of course, even now, a fairly big proportion still don't have AU on, since they never installed SP2 because they never update their system... )
RTFA: the only reason that this month was different was a bug in Skype's network allocation algorithm, a bug of a type that was triggered by a normal Patch Tuesday mass rebooting event.
Even if the EU & DoJ didn't exist, Microsoft is what it is today in a large part through the use of backscratch ("I'll scratch your back if you scratch mine") exclusivity agreements with all the major OEMs. MS introducing its own hardware would be regarded by the OEMs as a severe breach of "trust" (for want of a better word), and MS does not want the CEO of Sony to start wondering whether the time of Linux on the Desktop has come...
Which is all actually rather a pity, since Microsoft hardware (mice, keyboards, XBoxes etc.) is actually pretty good, and the retail PC market could well to with a bit of operating system diversification; but such is life...
Whilst the situation is a bit better with major OEMs (e.g. you don't tend to get stuff like this, it's not much. The Vista Certified etc. stickers are not saying that that *particular* hardware configuration has been tested by MS (or, indeed, anyone), all they're saying is that the specification meets the recommended MS guidelines for Vista. Because -- here's a key point -- the actual hardware configuration will change regularly. A Sony ordered on Tuesday might have a no-name-brand A motherboard, whilst the identical machine ordered on Thursday might have a no-name-brand B one. The same spec, of course, since with the big-name-brands the actual motherboard manufacturer is likely making that entire line of motherboards specifically for Sony, to Sony's specifications; but the actual manufacturer and hardware is definitely not fixed. An analogy might be KFC: the product you see is pretty homogenised, but the chickens do not all come from one vast, centralised farm; they're bought from many hundreds of seperate farms, and the farm that the chicken you're eating will have come from will vary from week to week.
All this so far is generalised Microsoft/Linux model vs Apple model, rather than specific to Vista problems, but frankly most people here seem to have short memories: XP experienced vast amounts of hardware compatibility problems when it was first introduced, as did 2000, NT (especially), 98, 95, 3.1, etc. (I don't knw whether anyone's been able to tell whether ME did, since it had so many problems: who could tell which were due to hardware compatibility..?). Largely, this will get better with time (over a timescale of years, not months) -- part of the problem, I think, is that a lot of manufacturers are just using the same identical drivers they designed for XP and expecting them to work; and they often do, but sometimes don't (as happened with 2000 to XP, 95 to 98, etc.).
Apple makes the hardware that their OS will run on, and thus the device drivers for that hardware. Microsoft doesn't.
I'm afraid Wonko's right. The total efficiency converges to Carnot's. If there is any *usable* waste heat left at the end of a cycle to put into another heat engine, then the first cycle wasn't running at full (reversible) Carnot efficiency.
BTW, if a heat engine were ever to be used in a closed system, then its efficiency would quickly converge to 0%, since the hot source would cool down and the cold source would heat up! The Carnot cycle, as I said in my other post, assumes infinite, constant temperature hot and cold sources, i.e. effectively the same as a system where heat is constantly added to the hot source (by a nuclear reactor) and taken away from the cold source (by a running river) to maintain a constant temperature.
Tancredo, Brownback, and Huckabee.
(Grammar Nazi side-note: no apostrophe needed for the plural candidates).
Or you could RTFA and realise that it's talking about quantum tunnelling. Just an idea.
In case you don't know, the two postulates are:
First postulate - The laws of physics are the same in all inertial frames of reference. In other words, there are no privileged inertial frames of reference.
Second postulate - The speed of light in a vacuum is a constant, which is independent of the motion of the light source.
So, which one of them do you think is wrong? If you're right, one of them must be...
5/7 of the speed of bad news, but only half as fast as the speed of rumours.
- It exists. Yup.
- It has volume. Nope. It has one measureable spatial quality, which is wavelength. To ask what the other two dimensions of a photon are ('volume' implies three) is meaningless. Remember, a photon is a packet of energy, a propogating electromagnetic wave (the electric and magnetic components at right angls to each other). It has no well-defined volume.
- It is "detectable as a particle/atom or whatever". Ummm... Oh dear. A particle, in some sense, certainly: google "wave-particle duality". An atom, no.
To get back to the issue at hand, the thing which might be confusing you is that is has no rest-mass. If it *did* have rest-mass, it couldn't move at the speed of light. But it certainly does have mass in the sense that it is moving, so has kinetic energy, and, as Einstein pointed out, energy and mass are just two words for the same thing. It's energy due to its momentum is given by E=h*frequency=hc/wavelength.
You're mixing up Newtonian with Relativistic equations. Try p=gamma*m_0*v, where m_0 is the rest mass, and gamma is 1/sqrt(1-v^2/c^2).
Caveat: you could, of course, arrange to, say, have the other party shoot their pet cat if their result comes up 'left', and repreive it if 'right'. By measuring your particle, you could then know whether the cat is dead or alive. But in this case, no information is actually transferred across the entanglement: the situation is the same as if you had carried a sealed envelope with you written by the other party, with whether they are planning to kill their cat written inside it, that you then open at a later time. You know whether the cat is dead or not, but again, information is not of course instantaneously transferred from the other party to you.