StCredZero "writes wildness. Fuji drew up the photocopying machine which automatically translates the document from English from Japanese. That is the clean nut. With respect to appearance, as for the copier the text, as for OCR what kind of section text, to send that to the translation engine, and in the place English". You reset, or can grasp.
Bull. The poor have no savings because they have no incentive to save -- their savings is destroyed by inflation.
Bull yourself. The interest on zero-risk FDIC-insured savings accounts beats inflation, as long as inflation is kept low, which is a job the Federal Reserve has been successful in performing for quite some time now. *Nobody's* savings are being "destoyed" by the inflation we have here in the US.
Also, hoarding money is *terrible* for the economy. Value in this world is created through trade, and the more trade there is the bigger and healthier the economy. If you have a bunch of money stashed under your mattress and you're not letting anyone use it, that money is not contributing to the economy. Putting that money to work increases the size of the economy.
I don't understand you people. Why do you care how much the dollar has depreciated since 1913? Have you been hoarding cash under your mattress since then? Do you have a time machine that brings 1913 dollars into the present? The value of the 1913 dollar has zero relevance to transactions going on today; it is not proof of any conspiracy to rob you of your value. The only thing that really matters is how inflation affects your assets, and if you invest your money, even just in a savings account at the bank, you'll beat inflation as long as the Fed is doing its job. You should be *supporting* the Fed in its drive to keep inflation reasonable.
The only valid point you guys have is that politicians probably shouldn't be trusted with managing the money supply, because they have a huge incentive to print far too much money and cause unchecked inflation. But that's why we *need* something like the Fed; something not immediately beholden to the will of whatever politicians are in office right now, to manage the money supply. We can't go back to the gold standard; that cat is out of the bag and the gold standard has its own set of problems anyhow. We just have to do our best to keep control of the money supply uncorrupted by politics.
Insightful, not funny! Well, a bit of both. But this truly is a lesson the entire software industry needs to learn. Where are my mod points when I need them...
Absolutely. I'd go so far as to say that Euphoria, or something like it, will be the next leap in gaming immersion, now that we've long passed the point of diminishing returns in graphics. You can't have truly convincing human characters in games without something like Euphoria. But first they will have to get it to do more than just make guys fall over in more varied ways. The other day I found some really interesting research into generating dynamic walk animations; something as simple as this (well it's not really simple, but you know what I mean) added to today's game engines would be a tremendous improvement. It just has to make its way from academia into the real world. NaturalMotion should be all over this guy.
They will probably continue to work on that, as they still want to sell Havok so they have to make a product that's attractive to their customers. However, when Intel's new graphics card comes out, expect Havok's support for it to be especially good.
There's no point to accelerating the core logic, but try to think outside the box. There could be wins in accelerating other tasks which would allow the core logic to make better decisions. For example, you could accelerate computing visibility to find better hiding places, or accelerate pathfinding to get better paths for the core logic to choose from, or accelerate procedural animation to allow for more varied and realistic behaviors. A product which did these things could credibly be marketed as an "AI accelerator", even though it doesn't accelerate finite state machines.
I think they're going to make it run on Larrabee, their in-development x86-based graphics card to compete with GeForce and Radeon. It's hard to imagine a more perfect match, actually.
The genius of the design is that the bits can be moved along the nanowires, allowing tens to hundreds of bits or maybe more to be accessed by only one reader. The readers can be fabricated in an array on a chip, and the wires can be hung from above, storing the data vertically. AFAIK they haven't yet gotten to the point of figuring out fabrication issues for the nanowire parts, like making a vertically oriented array and aligning them to readers. So far they have been working on getting the racetrack part working. That is, they have been working on using an electric current to shift magnetic domains longitudinally along a nanowire, and reading/writing the domains. And actually, the article seems to suggest that they are ignoring the 3-dimensional nanowire fabrication issues for now, and are going to make prototypes with the wires fabricated traditionally, 2-dimensionally, on a chip surface, which may still be competitive with Flash.
As for heat issues, Hopefully the amount of current necessary will be small and thus the wires themselves will generate little heat. I would imagine that this design would have fewer transistors than, say, a DRAM, since the transistors will not be storing the data themselves. The transistors remain 2-dimensional, only on the chip surface. The wires are the only 3-dimensional part.
The post you linked to is written by someone who clearly doesn't understand that sometimes data crosslinks to data
Your insinuation is clearly false. He traces the exact crosslinks you assert he doesn't understand. You miss his point, which is that the crosslinks he is complaining about are unnecessary cruft, unbefitting a proposed standard, as proved by comparison to ODF. Furthermore, he has many other complaints besides the crosslinking stuff. The OOXML "standard" is littered with such cruft, due to its heritage as a practically 1:1 dump of all the hacks and bad ideas Microsoft has ever crammed into their previously opaque binary format, with hardly any effort put into normalizing the format to solve any of the obvious problems, plus a whole *new* set of problems from the rush-job XML translation.
The point is that the complete read-write implementation of OOXML, interoperable with Office, is clearly impossible in practice without access to the legacy Office codebase, and even a partial implementation is far harder than it has any right to be, proven by the example of ODF which is demonstrably easier. In fact, a partial implementation may not even be realistically possible because of all the interdependencies which ODF has fewer of. What is a standard if it is not interoperable; not implementable? What's the point of making it a standard if it's not going to be fully implemented by anyone but Microsoft?
it won't help if the "consensus" among wikipedia authors is wrong.
In many cases, knowing the popular consensus on an issue is just as important as knowing the truth. Wikipedia is tremendously useful as long as one understands the difference.
I'll also point out that consensus is the best approximation to truth available to us as a society (flawed though it is); consensus is the foundation of both democracy and the scientific peer review process. If you found something better than consensus at determining truth, you'd have the basis for a government better than democracy and a scientific process better than peer review.
Municipal WiFi is a stupid idea. It's completely impossible to get the kind of blanket fast, reliable coverage people should be able to expect from their Internet connection. WiFi's range is just too short and buildings attenuate the signal. You'd have to put a router every 50 feet over the entire city, including inside private property! Furthermore, mesh networking sucks; always has and always will. It multiplies the packet loss and latency and divides the bandwidth for each hop you go through.
Municipal WiMax, on the other hand, makes tons of sense. I love the idea of a public utility wireless broadband service to provide some competition for lazy telecoms; WiFi just isn't the way to do it. WiMax all the way. Use things for their intended purpose; that's my motto.
I'm sure it would be difficult to mold this chip into a GPU, but I'd like to point out that NVidia's GeForce 8 series actually does bother (a bit) with integer arithmetic, and is actually a scalar architecture, with no 4-way vector instructions (though it does pair multiple functional units with one instruction decoder, each functional unit executes a different thread).
The biggest differences between Sun's chip and the 8 series are probably in the memory architecture. The 8 series has a ginormous memory bandwidth and many specialized ways to access it, all wired up in a highly optimized pattern: z-buffering hardware, framebuffer blending hardware, many read-only texture sampling units (which are powerful processors in their own right with dedicated caches), and local programmer-managed read-only and read-write memories for each core.
Who says you're going to listen to the ads? A Google phone would have a large screen, Internet, GPS, and Google maps. Imagine the absolute gold mine that location-based advertising could be on such a device. Personally, I think the potential of this market is at *least* as big as either of Google's current big markets, web search ads and targeted website ads (both of which, I should point out, would also be expanded by a Google Internet phone).
You forgot "encrypted". IMHO *every* connection should be encrypted. The reason we don't do it today is the certificate management problem. SSL solves the problem with centrally-managed certificate authorities, but that's too much management overhead for use on the entire Internet. We could bypass the certificate management problem by simply doing encryption like SSH does: the first time you connect to someone you cache their certificate and use it for all subsequent connections. You can't be sure that this initial certificate is correct, because there's no truly secure channel to verify it. However, you *can* be sure that you are always talking to the same person, which is better than today's Internet where there's no verification of the endpoints at all.
Here's how it would work in the real world: at home, you visit Google on your laptop. You trust your home network, so you get Google's correct cert and your communication is encrypted. Later, you go to Starbucks and use the insecure wireless network. An attacker tries to hijack your connection to Google, but his cert doesn't match Google's so your browser notifies you and the attack is thwarted.
What if you get the wrong cert initially? You go with your brand new laptop to Starbucks, visit Google for the first time and get the attacker's cert. The attacker intercepts all your Google searches for that day. That's bad. BUT, later, when you go home, you visit Google and your browser notifies you that the certs don't match. Now you *know* that your earlier connection was attacked, and you can fix it so you're not attacked in the future, which is far superior than today's situation, where you would never know and would remain vulnerable.
In conclusion, even though an SSH-like scheme of caching certificates on first connection can't prevent all attacks, it would provide a much higher level of routine security for the Internet compared to today's lack of security.
I disagree with your assertions. The generic solution is possible and a generic solution is always preferable. CFS is much better than mediocre, and most people will be content (and probably even happy) with it. If you disagree, show me the benchmarks where CFS performs significantly worse than another scheduler, and the difference is larger than the measurement error, and it's not easily fixed within days by Ingo.
Furthermore, CFS is not mandated; people are still free to patch their kernels. Nothing can be mandated in Linux, it's open source! Distros almost always apply many patches to their kernels for issues similar to this.
One of the hardware benchmark sites should invest in one of those high-speed cameras and do some *real* testing of all the latency figures tossed around these days. With one of those cameras you could verify manufacturers' monitor latency numbers *and* benchmark real whole-system latency from mouse click to gun fire.
On an unrelated note, I agree with Linus that a pluggable scheduler is a terrible idea. It is possible for one scheduler to do well on a huge variety of workloads. Pluggable schedulers would just encourage development of schedulers that are 2% better than CFS at one thing and suck for everything else, which in the end benefits very few people because few people do only one thing all the time (and those people can apply kernel patches if they really need that 2%). A desktop workload is really more demanding than any server workload because of the variety of tasks. In many cases desktop users need throughput just as much as server guys do. And kernel developers *do* care about desktop Linux, because as Linus pointed out they use it themselves, and besides, it's a more interesting problem.
So is the moral of the story is "Let them pirate your merchandise or they might use the competitions"?
In fact, yes, if you are a monopoly.
Let me tell you about the economic principle of "price discrimination". You make the most profit when you charge each person who buys your product pays the maximum price they would have been willing to pay for it. If someone isn't willing to pay for your product, then you don't lose any revenue by giving it to them for free. And if it costs you nothing to give it to them, *plus* it strengthens your monopoly position, thereby increasing your ability to extract cash from the people who *are* paying you, then it starts to look like a good idea. Also factor in the fact that people who pirate Windows will likely have to become paying locked-in Windows customers in the future when enforcement is stepped up in their countries; well then it's practically a no-brainer.
You should vote for Ron Paul even if you disagree with him (as I do) on abortion. Why? Because as president he wouldn't have the power to ban abortion, but he *would* have the power to change our foreign policy. His election as president (or just Republican party candidate, or even just serious contender in the primaries) would send a giant message to Washington (and in particular the Republican party) favoring his libertarian ideals of small government and less regulation, because that is where he differs with other candidates. It wouldn't really send much of a message about abortion.
If C|net really thinks that removing trivial user-interface functionality like bookmarks and history is going to significantly reduce Firefox's memory footprint or CPU usage, then I would suggest that C|net is not qualified to be giving advice to the Mozilla team on this subject. This guy is C|net's "expert on digital music and portable media"; has he ever even written a line of code?
Let's look at his other suggestions. Removing tabs would probably result in people opening fewer pages at a time, but people are already free to ignore tabs if they don't want to use them. There is no point in removing the functionality. (In fact, I would be willing to bet that one window with three tabs uses less memory than three windows). The same goes for extensions; people are free to not install any and removing the functionality would likely not further reduce the memory footprint.
Yup, basically, this guy has no idea what causes memory usage in Firefox. I'm glad that the Mozilla team will undoubtedly ignore his misguided advice. Here's a hint: the main driver of Firefox memory and CPU use is web pages. Parsing, rendering, and running scripts. Web pages are huge nowadays, with tons of scripting, huge images, and even videos, and all that stuff has to be kept in memory while you have a page open. If you want to make Firefox more efficient, don't look at the UI. Look at Gecko. Unfortunately, this means you have to be a programmer to make informed comments about Firefox's memory use.
As mr_matticus has been pointing out elsewhere on this thread, AT&T's 3G network basically sucks, and relatively few people are covered by it. Apple has known this all along, and in that light their decision makes slightly more sense, though I still think it was a mistake.
I guess Apple believes that visual voicemail and the ability to sell iPhones at retail with unsubsidized prices were worth the price of an exclusive contract with the least Internet-friendly carrier, because as far as I can tell those are the only two things Apple got out of this deal. When Steve Jobs was crowing about the fact that they got AT&T to make concessions, I was hoping for free VoIP over WiFi, free instant messaging instead of 10c per text message, an open development platform, no 2-year contracts; you know, things which would reduce the carriers' stranglehold and increase competition; a revolution like iTunes brought to the music market. Visual Voicemail is nice, but it's not really a "concession" on AT&T's part.
What you would *like* to say is that AT&T's network sucks, therefore lack of HSDPA is a stupid reason to avoid the iPhone. What you *have* been saying is that a lack of 3G is a stupid reason to avoid the iPhone. It's subtle, I know, but I hope you can see the difference. Anyway, I'm done arguing about that.
10% of the population is not 10% of the country.
It's the 10% most active cell phone using areas in the country, which is far more useful than choosing 10% of the country at random. Let me point out that population is the best available metric for measuring cell phone coverage. It is folly to measure cell coverage in units of area, because cell service is provided to people, not land. And even when you travel you go to places where people are; unless you're trying to get away from people and in that case complaining about lack of cell service is a bit disingenuous...
The speeds and connectivity are not substantially better than GPRS
Well that's certainly interesting anecdotal evidence. My 3G EV-DO experience has been great with both Verizon and Sprint; I've never used HSDPA, and all I've heard about it before was that it was actually *faster* than EV-DO, because it has better theoretical bandwidth and latency. If it's really as bad as you say in the real world, that makes the argument you'd like to make a lot stronger (and makes me wish for a CDMA iPhone a lot more, because I'm not giving up 3G even for Apple). I hope you won't take offense, however, when I say that I always take my anecdotal evidence with a grain of salt (especially here on Slashdot!).
The US missed the boat.
Cingular did for sure, but Verizon and Sprint just barely made it. Actually, Sprint's now ready to board the next one; they are rolling out WiMAX as their 4G service starting *this year*. The US is finally catching up. We can only hope that wireless Internet will soon be enough competition to break the DSL/Cable duopoly so we can board the 100Mbps full-duplex FTTH boat too...
I'm afraid it's you who isn't being logical, and as evidence I present the above statement and rest my case. Lack of 3G is a problem for the iPhone, full stop. The iPhone is competing in the market with other phones that do have 3G. There are various and sundry causes for the lack of 3G in the iPhone, some of which have to do with the hardware, and some of which have to do with AT&T's network, and none of which change the fact that the iPhone would be a much more compelling purchase if you could browse the web with 3G as you can on competing phones.
If the iPhone had 3G, you wouldn't be able to use it in 90% of the country
Firstly, that figure is wrong. Cingular had HSDPA to over 10% of the population two years ago, and coverage has improved since then. Secondly, even if it were true, 10% of the country (and a much larger percentage of iPhone customers, living in urban tech centers likely to be covered) would still be completely justified in wanting an HSDPA iPhone now.
Not buying something because it doesn't have a feature that "will be" implemented in the future, to some unknown extent and performance level, is absurd.
There is no question that the iPhone will eventually have 3G. There is no question that AT&T will expand their coverage (~60 "markets" within the year according to them). There is no question that HSDPA's performance is light-years beyond EDGE when available. It is not at all absurd to put off spending $600 now if you have a reasonable belief that a year from now you can get something far better.
You seem to be saying "If I can't use HSDPA in XX% of the US right now, I don't want it", which is absurd. A reasonable person would say "If I have HSDPA coverage in my home area, or if I expect it in the next year or two, I shouldn't spend $600 now on a phone without HSDPA". And a quite large percentage (more than 10%, likely approaching 50%) of iPhone customers can say that. Furthermore, nearly 100% of iPhone customers can say "The iPhone doesn't have 3G but [for example] the Motorola Q does", which makes the iPhone significantly less attractive if internet access is something you value.
Slashdot Mirror
StCredZero "writes wildness. Fuji drew up the photocopying machine which automatically translates the document from English from Japanese. That is the clean nut. With respect to appearance, as for the copier the text, as for OCR what kind of section text, to send that to the translation engine, and in the place English". You reset, or can grasp.
Bull yourself. The interest on zero-risk FDIC-insured savings accounts beats inflation, as long as inflation is kept low, which is a job the Federal Reserve has been successful in performing for quite some time now. *Nobody's* savings are being "destoyed" by the inflation we have here in the US.
Also, hoarding money is *terrible* for the economy. Value in this world is created through trade, and the more trade there is the bigger and healthier the economy. If you have a bunch of money stashed under your mattress and you're not letting anyone use it, that money is not contributing to the economy. Putting that money to work increases the size of the economy.
I don't understand you people. Why do you care how much the dollar has depreciated since 1913? Have you been hoarding cash under your mattress since then? Do you have a time machine that brings 1913 dollars into the present? The value of the 1913 dollar has zero relevance to transactions going on today; it is not proof of any conspiracy to rob you of your value. The only thing that really matters is how inflation affects your assets, and if you invest your money, even just in a savings account at the bank, you'll beat inflation as long as the Fed is doing its job. You should be *supporting* the Fed in its drive to keep inflation reasonable.
The only valid point you guys have is that politicians probably shouldn't be trusted with managing the money supply, because they have a huge incentive to print far too much money and cause unchecked inflation. But that's why we *need* something like the Fed; something not immediately beholden to the will of whatever politicians are in office right now, to manage the money supply. We can't go back to the gold standard; that cat is out of the bag and the gold standard has its own set of problems anyhow. We just have to do our best to keep control of the money supply uncorrupted by politics.
Insightful, not funny! Well, a bit of both. But this truly is a lesson the entire software industry needs to learn. Where are my mod points when I need them...
Absolutely. I'd go so far as to say that Euphoria, or something like it, will be the next leap in gaming immersion, now that we've long passed the point of diminishing returns in graphics. You can't have truly convincing human characters in games without something like Euphoria. But first they will have to get it to do more than just make guys fall over in more varied ways. The other day I found some really interesting research into generating dynamic walk animations; something as simple as this (well it's not really simple, but you know what I mean) added to today's game engines would be a tremendous improvement. It just has to make its way from academia into the real world. NaturalMotion should be all over this guy.
They will probably continue to work on that, as they still want to sell Havok so they have to make a product that's attractive to their customers. However, when Intel's new graphics card comes out, expect Havok's support for it to be especially good.
There's no point to accelerating the core logic, but try to think outside the box. There could be wins in accelerating other tasks which would allow the core logic to make better decisions. For example, you could accelerate computing visibility to find better hiding places, or accelerate pathfinding to get better paths for the core logic to choose from, or accelerate procedural animation to allow for more varied and realistic behaviors. A product which did these things could credibly be marketed as an "AI accelerator", even though it doesn't accelerate finite state machines.
I think they're going to make it run on Larrabee, their in-development x86-based graphics card to compete with GeForce and Radeon. It's hard to imagine a more perfect match, actually.
Here's the concept, with a nice animated gif: http://www.almaden.ibm.com/spinaps/research/sd/?racetrack
The genius of the design is that the bits can be moved along the nanowires, allowing tens to hundreds of bits or maybe more to be accessed by only one reader. The readers can be fabricated in an array on a chip, and the wires can be hung from above, storing the data vertically. AFAIK they haven't yet gotten to the point of figuring out fabrication issues for the nanowire parts, like making a vertically oriented array and aligning them to readers. So far they have been working on getting the racetrack part working. That is, they have been working on using an electric current to shift magnetic domains longitudinally along a nanowire, and reading/writing the domains. And actually, the article seems to suggest that they are ignoring the 3-dimensional nanowire fabrication issues for now, and are going to make prototypes with the wires fabricated traditionally, 2-dimensionally, on a chip surface, which may still be competitive with Flash.
As for heat issues, Hopefully the amount of current necessary will be small and thus the wires themselves will generate little heat. I would imagine that this design would have fewer transistors than, say, a DRAM, since the transistors will not be storing the data themselves. The transistors remain 2-dimensional, only on the chip surface. The wires are the only 3-dimensional part.
The point is that the complete read-write implementation of OOXML, interoperable with Office, is clearly impossible in practice without access to the legacy Office codebase, and even a partial implementation is far harder than it has any right to be, proven by the example of ODF which is demonstrably easier. In fact, a partial implementation may not even be realistically possible because of all the interdependencies which ODF has fewer of. What is a standard if it is not interoperable; not implementable? What's the point of making it a standard if it's not going to be fully implemented by anyone but Microsoft?
The text zoom feature exaggerates this bug. They rewrote the rendering engine for Firefox 3, so it will hopefully be fixed.
All those people (you know who you are) who kept saying "I'd buy an iPhone without the phone", you better step up.
I'll also point out that consensus is the best approximation to truth available to us as a society (flawed though it is); consensus is the foundation of both democracy and the scientific peer review process. If you found something better than consensus at determining truth, you'd have the basis for a government better than democracy and a scientific process better than peer review.
Municipal WiFi is a stupid idea. It's completely impossible to get the kind of blanket fast, reliable coverage people should be able to expect from their Internet connection. WiFi's range is just too short and buildings attenuate the signal. You'd have to put a router every 50 feet over the entire city, including inside private property! Furthermore, mesh networking sucks; always has and always will. It multiplies the packet loss and latency and divides the bandwidth for each hop you go through.
Municipal WiMax, on the other hand, makes tons of sense. I love the idea of a public utility wireless broadband service to provide some competition for lazy telecoms; WiFi just isn't the way to do it. WiMax all the way. Use things for their intended purpose; that's my motto.
I'm sure it would be difficult to mold this chip into a GPU, but I'd like to point out that NVidia's GeForce 8 series actually does bother (a bit) with integer arithmetic, and is actually a scalar architecture, with no 4-way vector instructions (though it does pair multiple functional units with one instruction decoder, each functional unit executes a different thread).
The biggest differences between Sun's chip and the 8 series are probably in the memory architecture. The 8 series has a ginormous memory bandwidth and many specialized ways to access it, all wired up in a highly optimized pattern: z-buffering hardware, framebuffer blending hardware, many read-only texture sampling units (which are powerful processors in their own right with dedicated caches), and local programmer-managed read-only and read-write memories for each core.
Who says you're going to listen to the ads? A Google phone would have a large screen, Internet, GPS, and Google maps. Imagine the absolute gold mine that location-based advertising could be on such a device. Personally, I think the potential of this market is at *least* as big as either of Google's current big markets, web search ads and targeted website ads (both of which, I should point out, would also be expanded by a Google Internet phone).
You forgot "encrypted". IMHO *every* connection should be encrypted. The reason we don't do it today is the certificate management problem. SSL solves the problem with centrally-managed certificate authorities, but that's too much management overhead for use on the entire Internet. We could bypass the certificate management problem by simply doing encryption like SSH does: the first time you connect to someone you cache their certificate and use it for all subsequent connections. You can't be sure that this initial certificate is correct, because there's no truly secure channel to verify it. However, you *can* be sure that you are always talking to the same person, which is better than today's Internet where there's no verification of the endpoints at all.
Here's how it would work in the real world: at home, you visit Google on your laptop. You trust your home network, so you get Google's correct cert and your communication is encrypted. Later, you go to Starbucks and use the insecure wireless network. An attacker tries to hijack your connection to Google, but his cert doesn't match Google's so your browser notifies you and the attack is thwarted.
What if you get the wrong cert initially? You go with your brand new laptop to Starbucks, visit Google for the first time and get the attacker's cert. The attacker intercepts all your Google searches for that day. That's bad. BUT, later, when you go home, you visit Google and your browser notifies you that the certs don't match. Now you *know* that your earlier connection was attacked, and you can fix it so you're not attacked in the future, which is far superior than today's situation, where you would never know and would remain vulnerable.
In conclusion, even though an SSH-like scheme of caching certificates on first connection can't prevent all attacks, it would provide a much higher level of routine security for the Internet compared to today's lack of security.
I disagree with your assertions. The generic solution is possible and a generic solution is always preferable. CFS is much better than mediocre, and most people will be content (and probably even happy) with it. If you disagree, show me the benchmarks where CFS performs significantly worse than another scheduler, and the difference is larger than the measurement error, and it's not easily fixed within days by Ingo.
Furthermore, CFS is not mandated; people are still free to patch their kernels. Nothing can be mandated in Linux, it's open source! Distros almost always apply many patches to their kernels for issues similar to this.
One of the hardware benchmark sites should invest in one of those high-speed cameras and do some *real* testing of all the latency figures tossed around these days. With one of those cameras you could verify manufacturers' monitor latency numbers *and* benchmark real whole-system latency from mouse click to gun fire.
On an unrelated note, I agree with Linus that a pluggable scheduler is a terrible idea. It is possible for one scheduler to do well on a huge variety of workloads. Pluggable schedulers would just encourage development of schedulers that are 2% better than CFS at one thing and suck for everything else, which in the end benefits very few people because few people do only one thing all the time (and those people can apply kernel patches if they really need that 2%). A desktop workload is really more demanding than any server workload because of the variety of tasks. In many cases desktop users need throughput just as much as server guys do. And kernel developers *do* care about desktop Linux, because as Linus pointed out they use it themselves, and besides, it's a more interesting problem.
Let me tell you about the economic principle of "price discrimination". You make the most profit when you charge each person who buys your product pays the maximum price they would have been willing to pay for it. If someone isn't willing to pay for your product, then you don't lose any revenue by giving it to them for free. And if it costs you nothing to give it to them, *plus* it strengthens your monopoly position, thereby increasing your ability to extract cash from the people who *are* paying you, then it starts to look like a good idea. Also factor in the fact that people who pirate Windows will likely have to become paying locked-in Windows customers in the future when enforcement is stepped up in their countries; well then it's practically a no-brainer.
You should vote for Ron Paul even if you disagree with him (as I do) on abortion. Why? Because as president he wouldn't have the power to ban abortion, but he *would* have the power to change our foreign policy. His election as president (or just Republican party candidate, or even just serious contender in the primaries) would send a giant message to Washington (and in particular the Republican party) favoring his libertarian ideals of small government and less regulation, because that is where he differs with other candidates. It wouldn't really send much of a message about abortion.
Seriously, WTF?
If C|net really thinks that removing trivial user-interface functionality like bookmarks and history is going to significantly reduce Firefox's memory footprint or CPU usage, then I would suggest that C|net is not qualified to be giving advice to the Mozilla team on this subject. This guy is C|net's "expert on digital music and portable media"; has he ever even written a line of code?
Let's look at his other suggestions. Removing tabs would probably result in people opening fewer pages at a time, but people are already free to ignore tabs if they don't want to use them. There is no point in removing the functionality. (In fact, I would be willing to bet that one window with three tabs uses less memory than three windows). The same goes for extensions; people are free to not install any and removing the functionality would likely not further reduce the memory footprint.
Yup, basically, this guy has no idea what causes memory usage in Firefox. I'm glad that the Mozilla team will undoubtedly ignore his misguided advice. Here's a hint: the main driver of Firefox memory and CPU use is web pages. Parsing, rendering, and running scripts. Web pages are huge nowadays, with tons of scripting, huge images, and even videos, and all that stuff has to be kept in memory while you have a page open. If you want to make Firefox more efficient, don't look at the UI. Look at Gecko. Unfortunately, this means you have to be a programmer to make informed comments about Firefox's memory use.
As mr_matticus has been pointing out elsewhere on this thread, AT&T's 3G network basically sucks, and relatively few people are covered by it. Apple has known this all along, and in that light their decision makes slightly more sense, though I still think it was a mistake.
I guess Apple believes that visual voicemail and the ability to sell iPhones at retail with unsubsidized prices were worth the price of an exclusive contract with the least Internet-friendly carrier, because as far as I can tell those are the only two things Apple got out of this deal. When Steve Jobs was crowing about the fact that they got AT&T to make concessions, I was hoping for free VoIP over WiFi, free instant messaging instead of 10c per text message, an open development platform, no 2-year contracts; you know, things which would reduce the carriers' stranglehold and increase competition; a revolution like iTunes brought to the music market. Visual Voicemail is nice, but it's not really a "concession" on AT&T's part.
You seem to be saying "If I can't use HSDPA in XX% of the US right now, I don't want it", which is absurd. A reasonable person would say "If I have HSDPA coverage in my home area, or if I expect it in the next year or two, I shouldn't spend $600 now on a phone without HSDPA". And a quite large percentage (more than 10%, likely approaching 50%) of iPhone customers can say that. Furthermore, nearly 100% of iPhone customers can say "The iPhone doesn't have 3G but [for example] the Motorola Q does", which makes the iPhone significantly less attractive if internet access is something you value.