The US army includes a load of good folks (and a much smaller number of bad ones). The soldiers are not the problem, their superiors are.
In my current job, I've met and worked with several of those superiors. These are NOT dumb people; not by a long shot. In fact, I'd call your average American Army General a rather brilliant, engaging person to talk to. And while I've heard rumor from civilians of a stereotypical "grunt" in the armed forces who does what he's told like an unthinking automaton and can't reason his way out of a paper bag, I've never met one of these mythical people in the actual armed forces.
In any case, if you don't like what the Army's current strategic goals are, you need look no further than the Commander In Chief. Everyone else is just trying to accomplish his goals as best they can. Although you'll never get another chance to vote him out of office, you can certainly decrease the power of those who think like him by getting out and voting against Republicans this year and in 2008.
But we're getting off-topic. The real issue is whether TPM contains anything that might be useful to an army in encryption and identification. And I'd argue it undeniably does. Sure, there are downsides to TPM, but this is one case where the module might be honestly handy to have.
It's too bad your comment was worded in such an inflammatory nature, because your point is right-on. This version of the GPLv3 does absolutely nothing to address any of the serious problems the first one had with regards to making GPLv3 software commercially inviable. It will guarantee that GPLv2 will never die (since a lot of publishers will not want to give up the rights they had under GPLv2), and thus be Yet Another Open Source License, of which we have enough already, thank you.
In other words, GPLv3 is pointless work done only for Stallman's personal satisfaction with no practical value.
I don't think too many people here would really understand that..
I think more here would than anywhere. But add to that the increased expense of writing and debugging good multithreaded solutions, the decreased predictability of time slot allocation, etc., it's definitely no panacea. 8 cores "never" being useful is a strong statement, but most today would benefit more from the die space and engineering resources being spent elsewhere.
Am I the only one who found the Infocom game "A Mind Forever Voyaging" highly emotional? I'm not sure that I cried, but it certainly made me feel like wanting to. It was, of course, a text adventure from Infocom with no graphics whatsoever, which I played on an IBM XT. It was definitely made more emotional because of the interactivity.
There's also been plenty of tears shed in MMORPGs like Puzzle Pirates and kin, as alliances are created and destroyed. In that case, I suppose it's the other players making the folks emotional, with the game merely the conduit.
Isn't this the same thing they said about 64bit chips?
Good point... yes, Intel said this about 64bit chips, and they were right. Almost nobody needs 64bit chips. But now virtually all chips are 64bits, wasting a lot of die real estate and engineering effort because of the perceived benefits driven more by AMD's marketing than reality. It's quite possible 8 cores could end up in the same boat-- AMD pushing it for no valid technological reason and Intel being forced to follow suit.
I wonder if Apple will ever bring another really innovative computer to market? It hasn't done so since the Apple II and the orignal Mac lines of PCs.
Actually, Macs have been innovation leaders pretty much since they were released up through today. They drove widespread adoption of everything from CD-ROMs to USB to wireless networking. It's true they traditionally do not offer as much bang for the buck in terms of processor or polygon-count rendering, and tended to suffer on benchmarks that time those specific items. But in terms of innovation, it's hard to argue that any other platform has brought as much to the table as the Mac has over the years, from the original Mac to the Mac ][ to the various software innovations in the 90's, and through to the iMacs, they've always provided a lot of utility and forward-thinking for the money. I'm kind of of the opinion that Linux is only "free" if your time is worthless, and a Windows install is reasonable to start out with (as long as you don't want to do any of the things in the software packages the Mac ships with) but tends to get "bit rot" awfully fast. The Macs tend to stay clean and usable throughout their life. Obviously if you can't afford them that's certainly unfortunate, but you can't really argue Apple isn't innovative.
And my iPod works great, too.
You've obviously got some strange emotional feelings against what is, after all, just a company with some good products.
Plenty of folks make Watt-friendly integrated graphics chipsets for laptops, including Intel. They just get laughed at by gamers. When it comes to graphics cards, the market still prefers performance over power consumption, and that's probably not going to change too much anytime soon. Unlike the more complex instruction sets in PCs, I doubt graphics cards have a lot of optimization wiggle-room when it comes to eeking out more performance per Watt. So you're pretty much left to die-shrinking, for which the outlays can be expensive.
In general, MORE TIME TESTING = MORE STABLE PRODUCT.
Have you really found this to be true? To me it seems like the projects that take the longest to test, especially when the testing mostly happens at the end of the development cycle, are the ones for which the design was the most muddled-- and no amount of testing can compensate for that. I'd guess that your typical Windows release gets a whole heck of a lot more testing than your typical Linux release, for example.
It really comes down to memory. Both AMD and Intel have decent dual-core designs now. AMD's has an integrated memory bus, which speeds accesses to main RAM, while Intel has a unified cache, which allows reallocation of the cache much more flexibly. I suspect it's the unified cache that lets the Core Duo and Core 2 Duo clean AMD's clock when it comes to single-threaded performance... the one thread gets almost all the cache. On AMD's design, it can't have more than 50%.
As for multithreaded performance, the Core 2 Duo is still no slouch, though. I don't think AMD's going to be wiping any floors with Intel any day soon.
Traditionally the United States bought more oil from Venezuela than any middle eastern country. Of course, now that we are trying to undermine their (democratically elected) government, we've shifted some of that buying to our (authoritarian) middle eastern allies.
Objective-C was a pretty nice language, but it lost. Now it's pigeon-holing Apple into a tiny subset of the world's developers, who are becoming increasingly fanatical. If they really want broad-based industry and developer support, they will have to dump it eventually. For now, they're getting by with bridges to a lot of the more popular scripting-style languages, but the fact that the Java bridge failed is proof that this approach can't work for "real" languages.
I'm a bit of a heretic, but I think Apple might do well to wholeheartedly embrace.NET/C#. I'll bet they could get great licensing deals from Microsoft, giving them a jump-start, in exchange for rapidly increasing cross-platform interest in the language over Java.
Your HD-DVD vs. Blu-Ray "common knowledge" links are amazingly inaccurate. ("Common knowledge" is very often not "correct knowledge".) Blu-Ray and HD-DVD support *exactly* the same codecs, so image and sound quality is going to be *exactly* the same. Blu-Ray has a lot more content coming out for it, since twice as many studios are producing for it. And the PS3 will be cheaper than most HD-DVD players. By the end of this Christmas shopping season, I'd be surprised if HD-DVD has anyone still believing in its long-term viability. Articles like the ones you cited, of course, are largely fueled by Microsoft money/fandom and anti-Sony propaganda, not reality.
When I was young I thought my adult life, if it happened, would be spent in a nuclear post-apocalyptic wasteland, so I suppose you can be forgiven for falling for the latest fear-mongering hype.
"But, keep in mind that NTFS remains proprietary and Microsoft can break it for newly written files any time it suits their business purposes to do so. All it takes is one update."
This is pure FUD. The entire reason Microsoft is the most successful business in the world is that they stay compatible with more stuff for longer than anyone else. (In many cases in practice, it's the other folks staying compatible with them, but it's the same issue.) In any case, they're not going to go and break forwards compatibility with previous versions for a long, long time.
There are lots of reasons to dislike Microsoft and distrust their software, but this isn't one of them.
Is this intended as flamebait? Performance in Java is as important as performance anywhere else. Most of the performance in any complex system depends more on the algorithms and how easy it is to debug a complex, fast algorithm, than on any inherent speed advantage of a particular language. I'd say the question of what design of network I/O and threading maximizes performance is at least as important a question in Java as any other language.
Please re-read my post starting from But even for 8bit math, the 16bit processor is going to be faster.... and realize that you can extend that 8:16 to be 32:64 and it still holds true. When doing 32bit math on a 64bit processor you have, effectively, twice as many registers.
I read it, I understood it and I disagree with it. You almost never have "twice as many registers" unless you do a LOT of hacky manipulation or you need to operate on numbers bigger than 4294967296. The 8:16 move was obvious because you almost always want to do math on numbers bigger than 256. 16->32 found that people operate on numbers bigger than 65536 often enough that it, too, always yielded significant performance increases. But unless you're loading registers with image data (usually better done through SIMD) or some other data stream, 64-bits is only useful for people who need to regularly deal with integers (note that floating point decimal numbers aren't affected) bigger than 2^32. This isn't very often.
Thus, encryption and naive image processing can benefit, but for most general purpose tasks an equivalent 64-bit processor will probably be slightly slower than its 32-bit counterpart. This is especially true if you consider where else those transistors could have been invested in the chip. x86 vs x86-64 has a slight exception to this because it was previously so hobbled and they took advantage of the ISA change to eliminate bogosities, but it's the exception not the rule.
And all of this applies to the rare instances when you need to deal with numbers bigger than 4294967296. Otherwise, the 64-bittedness is usually just adding extra data without providing any benefit. Other posters have pointed out why the x86 instruction set is a little bit of an exception since it was cleaned up a bit when it went to 64-bits, but in general 64-bits is usually a little slower than 32-bits for general purpose tasks.
We developed a rule of taking the engineers estimates and double them.
Doubling, tripling, or whatever is the naive approach to estimation. The problem with it is that while you'll probably be closer to reality the first couple times, especially with a young engineer or manager, you'll never improve your estimating ability. In order to improve your estimates, you have to try to make truly accurate estimates. If there is an overrun, there is something in the task list or the process itself you didn't estimate properly. (The last statement may seem obvious, but is often overlooked.)
Now, why does the "double what I'm told" technique tend to work? Because always "doubling" (or more) the estimate each time someone gives you one implies many layers of management and approval, which tends to increase the time (eg. if the original estimate required 2 "doublings", it implies 2 layers of management approval). So the original estimate should look at the org chart and take into account layers of approval required for each major decision, and use that as a multiplying factor (a deep vertical decision structure in a company acts as a multiplying factor because it applies to each major decision.)
The other obvious problem with this approach is that the estimation of a project is *not* the engineer's responsibility. It is the tech lead and/or manager's responsibility. The engineer has very little insight into bug rates, QA resource scheduling, meeting schedules, approval processes, etc. So if you sum the estimates for all features and double it, you're reneging on your responsibility for specific estimations and simply assuming you (as the manager) will add a 100% overhead to the project.
Unfortunately, there isn't a good alternative yet.
I think you meant "anymore". In the 90's there were quite a few "auction" sites on the internet. eBay's marketing and consolidation have driven most of them either offline or turned them into standard retailers. It's possible some new one could spring back up, but I think it's unlikely. More likely, I think, is eBay just fades into obscurity leaving only free sites like craigslist in its place as people get tired of the hassle and frustration of doing business through eBay.
You can get what is essentially a Mastercard with no annual fee and an APR somewhere between 13.5% and 22.5% depending on your credit rating, with no interest for 90 days after purchase. If you've got good credit, this pretty much amounts to the same thing you got. If not, good luck finding financing anyway.
Slashdot Mirror
The US army includes a load of good folks (and a much smaller number of bad ones). The soldiers are not the problem, their superiors are.
In my current job, I've met and worked with several of those superiors. These are NOT dumb people; not by a long shot. In fact, I'd call your average American Army General a rather brilliant, engaging person to talk to. And while I've heard rumor from civilians of a stereotypical "grunt" in the armed forces who does what he's told like an unthinking automaton and can't reason his way out of a paper bag, I've never met one of these mythical people in the actual armed forces.
In any case, if you don't like what the Army's current strategic goals are, you need look no further than the Commander In Chief. Everyone else is just trying to accomplish his goals as best they can. Although you'll never get another chance to vote him out of office, you can certainly decrease the power of those who think like him by getting out and voting against Republicans this year and in 2008.
But we're getting off-topic. The real issue is whether TPM contains anything that might be useful to an army in encryption and identification. And I'd argue it undeniably does. Sure, there are downsides to TPM, but this is one case where the module might be honestly handy to have.
It's too bad your comment was worded in such an inflammatory nature, because your point is right-on. This version of the GPLv3 does absolutely nothing to address any of the serious problems the first one had with regards to making GPLv3 software commercially inviable. It will guarantee that GPLv2 will never die (since a lot of publishers will not want to give up the rights they had under GPLv2), and thus be Yet Another Open Source License, of which we have enough already, thank you.
In other words, GPLv3 is pointless work done only for Stallman's personal satisfaction with no practical value.
I don't think too many people here would really understand that ..
I think more here would than anywhere. But add to that the increased expense of writing and debugging good multithreaded solutions, the decreased predictability of time slot allocation, etc., it's definitely no panacea. 8 cores "never" being useful is a strong statement, but most today would benefit more from the die space and engineering resources being spent elsewhere.
Am I the only one who found the Infocom game "A Mind Forever Voyaging" highly emotional? I'm not sure that I cried, but it certainly made me feel like wanting to. It was, of course, a text adventure from Infocom with no graphics whatsoever, which I played on an IBM XT. It was definitely made more emotional because of the interactivity.
There's also been plenty of tears shed in MMORPGs like Puzzle Pirates and kin, as alliances are created and destroyed. In that case, I suppose it's the other players making the folks emotional, with the game merely the conduit.
Isn't this the same thing they said about 64bit chips?
Good point... yes, Intel said this about 64bit chips, and they were right. Almost nobody needs 64bit chips. But now virtually all chips are 64bits, wasting a lot of die real estate and engineering effort because of the perceived benefits driven more by AMD's marketing than reality. It's quite possible 8 cores could end up in the same boat-- AMD pushing it for no valid technological reason and Intel being forced to follow suit.
I wonder if Apple will ever bring another really innovative computer to market? It hasn't done so since the Apple II and the orignal Mac lines of PCs.
Actually, Macs have been innovation leaders pretty much since they were released up through today. They drove widespread adoption of everything from CD-ROMs to USB to wireless networking. It's true they traditionally do not offer as much bang for the buck in terms of processor or polygon-count rendering, and tended to suffer on benchmarks that time those specific items. But in terms of innovation, it's hard to argue that any other platform has brought as much to the table as the Mac has over the years, from the original Mac to the Mac ][ to the various software innovations in the 90's, and through to the iMacs, they've always provided a lot of utility and forward-thinking for the money. I'm kind of of the opinion that Linux is only "free" if your time is worthless, and a Windows install is reasonable to start out with (as long as you don't want to do any of the things in the software packages the Mac ships with) but tends to get "bit rot" awfully fast. The Macs tend to stay clean and usable throughout their life. Obviously if you can't afford them that's certainly unfortunate, but you can't really argue Apple isn't innovative.
And my iPod works great, too.
You've obviously got some strange emotional feelings against what is, after all, just a company with some good products.
Plenty of folks make Watt-friendly integrated graphics chipsets for laptops, including Intel. They just get laughed at by gamers. When it comes to graphics cards, the market still prefers performance over power consumption, and that's probably not going to change too much anytime soon. Unlike the more complex instruction sets in PCs, I doubt graphics cards have a lot of optimization wiggle-room when it comes to eeking out more performance per Watt. So you're pretty much left to die-shrinking, for which the outlays can be expensive.
I very much doubt they will upgrade the macBookPro (as some suggest) befor they update the iMac, that remains their flagship product.
From last week's quarterly conference call:
"Apple sold 529,000 desktops during the quarter and 798,000 notebooks."
The ones God put in the Earth a few thousand years ago as a practical joke on those Copernican, old-Earth atheists.
In general, MORE TIME TESTING = MORE STABLE PRODUCT.
Have you really found this to be true? To me it seems like the projects that take the longest to test, especially when the testing mostly happens at the end of the development cycle, are the ones for which the design was the most muddled-- and no amount of testing can compensate for that. I'd guess that your typical Windows release gets a whole heck of a lot more testing than your typical Linux release, for example.
It really comes down to memory. Both AMD and Intel have decent dual-core designs now. AMD's has an integrated memory bus, which speeds accesses to main RAM, while Intel has a unified cache, which allows reallocation of the cache much more flexibly. I suspect it's the unified cache that lets the Core Duo and Core 2 Duo clean AMD's clock when it comes to single-threaded performance... the one thread gets almost all the cache. On AMD's design, it can't have more than 50%.
As for multithreaded performance, the Core 2 Duo is still no slouch, though. I don't think AMD's going to be wiping any floors with Intel any day soon.
Traditionally the United States bought more oil from Venezuela than any middle eastern country. Of course, now that we are trying to undermine their (democratically elected) government, we've shifted some of that buying to our (authoritarian) middle eastern allies.
Objective-C was a pretty nice language, but it lost. Now it's pigeon-holing Apple into a tiny subset of the world's developers, who are becoming increasingly fanatical. If they really want broad-based industry and developer support, they will have to dump it eventually. For now, they're getting by with bridges to a lot of the more popular scripting-style languages, but the fact that the Java bridge failed is proof that this approach can't work for "real" languages.
.NET/C#. I'll bet they could get great licensing deals from Microsoft, giving them a jump-start, in exchange for rapidly increasing cross-platform interest in the language over Java.
I'm a bit of a heretic, but I think Apple might do well to wholeheartedly embrace
Your HD-DVD vs. Blu-Ray "common knowledge" links are amazingly inaccurate. ("Common knowledge" is very often not "correct knowledge".) Blu-Ray and HD-DVD support *exactly* the same codecs, so image and sound quality is going to be *exactly* the same. Blu-Ray has a lot more content coming out for it, since twice as many studios are producing for it. And the PS3 will be cheaper than most HD-DVD players. By the end of this Christmas shopping season, I'd be surprised if HD-DVD has anyone still believing in its long-term viability. Articles like the ones you cited, of course, are largely fueled by Microsoft money/fandom and anti-Sony propaganda, not reality.
When I was young I thought my adult life, if it happened, would be spent in a nuclear post-apocalyptic wasteland, so I suppose you can be forgiven for falling for the latest fear-mongering hype.
"But, keep in mind that NTFS remains proprietary and Microsoft can break it for newly written files any time it suits their business purposes to do so. All it takes is one update."
This is pure FUD. The entire reason Microsoft is the most successful business in the world is that they stay compatible with more stuff for longer than anyone else. (In many cases in practice, it's the other folks staying compatible with them, but it's the same issue.) In any case, they're not going to go and break forwards compatibility with previous versions for a long, long time.
There are lots of reasons to dislike Microsoft and distrust their software, but this isn't one of them.
I'd be curious to see actual data that backs this up. Link?
>
Is this intended as flamebait? Performance in Java is as important as performance anywhere else. Most of the performance in any complex system depends more on the algorithms and how easy it is to debug a complex, fast algorithm, than on any inherent speed advantage of a particular language. I'd say the question of what design of network I/O and threading maximizes performance is at least as important a question in Java as any other language.
In Soviet Russia, the chemically assisted spatially controlled field tip evaporates YOU!
Please re-read my post starting from But even for 8bit math, the 16bit processor is going to be faster.... and realize that you can extend that 8:16 to be 32:64 and it still holds true. When doing 32bit math on a 64bit processor you have, effectively, twice as many registers.
I read it, I understood it and I disagree with it. You almost never have "twice as many registers" unless you do a LOT of hacky manipulation or you need to operate on numbers bigger than 4294967296. The 8:16 move was obvious because you almost always want to do math on numbers bigger than 256. 16->32 found that people operate on numbers bigger than 65536 often enough that it, too, always yielded significant performance increases. But unless you're loading registers with image data (usually better done through SIMD) or some other data stream, 64-bits is only useful for people who need to regularly deal with integers (note that floating point decimal numbers aren't affected) bigger than 2^32. This isn't very often.
Thus, encryption and naive image processing can benefit, but for most general purpose tasks an equivalent 64-bit processor will probably be slightly slower than its 32-bit counterpart. This is especially true if you consider where else those transistors could have been invested in the chip. x86 vs x86-64 has a slight exception to this because it was previously so hobbled and they took advantage of the ISA change to eliminate bogosities, but it's the exception not the rule.
And all of this applies to the rare instances when you need to deal with numbers bigger than 4294967296. Otherwise, the 64-bittedness is usually just adding extra data without providing any benefit. Other posters have pointed out why the x86 instruction set is a little bit of an exception since it was cleaned up a bit when it went to 64-bits, but in general 64-bits is usually a little slower than 32-bits for general purpose tasks.
We developed a rule of taking the engineers estimates and double them.
Doubling, tripling, or whatever is the naive approach to estimation. The problem with it is that while you'll probably be closer to reality the first couple times, especially with a young engineer or manager, you'll never improve your estimating ability. In order to improve your estimates, you have to try to make truly accurate estimates. If there is an overrun, there is something in the task list or the process itself you didn't estimate properly. (The last statement may seem obvious, but is often overlooked.)
Now, why does the "double what I'm told" technique tend to work? Because always "doubling" (or more) the estimate each time someone gives you one implies many layers of management and approval, which tends to increase the time (eg. if the original estimate required 2 "doublings", it implies 2 layers of management approval). So the original estimate should look at the org chart and take into account layers of approval required for each major decision, and use that as a multiplying factor (a deep vertical decision structure in a company acts as a multiplying factor because it applies to each major decision.)
The other obvious problem with this approach is that the estimation of a project is *not* the engineer's responsibility. It is the tech lead and/or manager's responsibility. The engineer has very little insight into bug rates, QA resource scheduling, meeting schedules, approval processes, etc. So if you sum the estimates for all features and double it, you're reneging on your responsibility for specific estimations and simply assuming you (as the manager) will add a 100% overhead to the project.
Unfortunately, there isn't a good alternative yet.
I think you meant "anymore". In the 90's there were quite a few "auction" sites on the internet. eBay's marketing and consolidation have driven most of them either offline or turned them into standard retailers. It's possible some new one could spring back up, but I think it's unlikely. More likely, I think, is eBay just fades into obscurity leaving only free sites like craigslist in its place as people get tired of the hassle and frustration of doing business through eBay.
Yeah, because it really costs $450 to reformat the hard drive and install the OS of your choice.
I guess you mean linux, because all other OSes are going to cost you quite a bit of money. Unless you're a thief.
Try checking out http://www.apple.com/financing/.
You can get what is essentially a Mastercard with no annual fee and an APR somewhere between 13.5% and 22.5% depending on your credit rating, with no interest for 90 days after purchase. If you've got good credit, this pretty much amounts to the same thing you got. If not, good luck finding financing anyway.