Mozilla is a hog, and if you're opening up a few documents, expect OpenOffice to use more too. I'd figure at least 100MB per user. Now we're talking 50GB of RAM.
I don't know what the author's been smoking, but I'd never put all the users on just one box. If a lab PC goes down, the user can switch to another. If your one and only Unix server goes down, everyone goes home. It would be better to split the processors between two smaller boxes, than to put everything in one.
The processors and memory are the largest part of the 4800's cost (especially in the configurations we're talking about. The chassis, backplane, and power supplies are relatively cheap.
And what the heck is going on with a SPARCstation 10 as the management console? Excuse me, but those have been discontinued for how many years? He mentions Office XP, so it's not as if this "report" was written in 1995. Sheesh.
I am a total Unix/Linux advocate, but this "report" is completely bogus.
Well, the guy was using a laptop, so that's not really an option, is it?
I've gone the build-it-yourself route. I selected the sound card, video, DVD, etc. Because it's a game system, I install Win98. Well, Diablo II runs like crap on it, other games are fine. I check the driver version, I check the message boards, I check Blizzard's site. No help.
As a lark, I install Windows NT 4.0. Diablo II now runs like a champ under that. Exact same hardware, same release of the drivers. But a big difference between 98 and NT 4.0. Why? who knows.
Bah. My next game machine will be a pre-built box . I don't want to deal with that crap any longer.
Heck, I may abandon Windoze altogether, and try to survive on Linux for gaming. Even then, I'll go with a pre-built box that has a proven configuration.
1. You get an e-mail virus written for your e-mail application. Your OS has multi-user security, so that the system binaries aren't affected. You log in as the root user, and clean up the affected files.
2. You get an e-mail virus. Since your OS doesn't have memory protection, it copies itself onto any executable it finds. You have to re-install your operating system because it's impractical to undo all the changes the virus made.
On the surface, this appears to be a terrible situation.
If you read a little further into the case, it doesn't seem like miscarriage of justice.
SBCCI and the like are non-profit corporations that earn money by publishing building standard codes. Cities then buy copies of the codes, and incorporate them into law, which is much easier than creating huge volumes of building codes by oneself.
You're allowed to look at the building codes for free; you can go to the city office. You can even make copies there. However, you're not allowed to publish copies the building code.
Do a web search to find out more about this case.
At any rate, that situation doesn't apply to what we're discussing with the W3C. The building codes themselves aren't proprietary... anyone is allowed to build a building according to the code.
People from the W3C have "acknowledged" that the Internet's growth has been due to open standards.
This isn't not even half the story. The Internet would not exist without open standards promoted by bodies like the IETF and W3C (until now).
In the 1980's and early 1990's there were a number of network protocols in use: DECNet, VINES, NETBIOS/NETBEUI (shudder), IPX/SPX, SNA, and more. None of them initially would have been as scalable as TCP/IP, however, if any of them had been truly open, it might have been possible to fix them.
But none of those other protocols were open... and where are they now? Nowhere.
It's the same situation for hypertext protocols. People and companies have proposed substantial improvments onto existing protocols. A notable example of this is Hyper-G, which was then commercialized by Hyperwave.com. It fixes a lot of problems with navigation, and stuff like broken links. However, there was never a free and open implementation, and so it has languished in obscurity for the last 5 years.
My question to the W3C is this: Do they have any evidence that proprietary protocols will foster continued growth of the Internet and the applications that run on top of it?
This is indeed worrisome, but some proliferation of licenses is necessary.
For example:
I was "around" (reading the mailing lists) when the Eiffel Forum License was created. Because of the nature of OO software development, it was felt that existing licenses (such as the LGPL) would not be appropriate. I don't recall all the discussions, but there were some good reasons why people didn't want to use the existing ones.
This is the "downside" to freedom; everyone tends to go off in their own direction. Of course M$ and the other industry players would be glad to consolidate end-user licenses down to just one. That would make things very simple... but not desirable or pleasant.
However, being in business myself, I know that a loss is a loss is a loss. It doesn't matter what you call it, it doesn't change the fact that there's more money going out the door than coming in.
Just because things look good on paper doesn't really mean that much. What really counts is how much money you have in the bank.
So, at any rate, it's good that I swiched to Debian last year.:-)
The trouble with stack machines is that they have a worse Von Neumann bottleneck than register machines. Everything has to go through the top of the stack.
Hmmmm.... I wonder about that. If you designed a CPU for Forth (or another stack-based language), you could have something where the first, say, 10 elements of the stack are in registers, and the rest in memory. As the stack grows, the bottom elements are moved off into main memory. Vice versa for shrinking. If done cleverly, there wouldn't be much need for copying data around.
Of course, I'm sure someone's already thought of this...
Well, you won't catch me spending a lot of time in front of a microwave transceiver.
Microwave ovens are optimized to heat water, but other types of microwaves (used in communications) can still do that too. Plus, I don't know what effect such high power levels would have (besides heating) but there might be some.
Good enough reason to be careful!
Re:Proper weapoin wont puncture the Aircraft
on
More On Tragedy
·
· Score: 2
The only problem with those "special guns that won't puncture steel" is that aircraft are mostly made out of aluminum these days. Better off with tazers, beanbags and such. Pepper spray would suck in such an enclosed environment.
As for hijackers intentionally crashing the plane they've taken over, it is perhaps not totally without precedent.
There was a hijacking of an aircraft out of India a few years ago. The guys were idiots, and forced the pilot to keep flying until the plane ran out of fuel.
Fortunately, the plane had just crossed over the shore, and the pilot tried to ditch in the ocean. Unfortunately, one of the hijackers grapped the flight yoke at the last second, and prevented the pilot from making a smooth water landing.
It probably didn't make as much news because only 1/3 of the people died.
At any rate, it won't as easy to do what they did again. I know that if I see someone trying to take over the plane I'm in, I'll do my best to immediately stop it. I'm not much of a fighter, but I'd rather take my chances with being shot a point blank range than having the plane I'm in crash into a building. At least I'll use up a bullet, and maybe give the guy behind me a chance...
The world got a lot colder and meaner Tuesday.
Re:What can be done about terrorism?
on
More On Tragedy
·
· Score: 5, Insightful
Every faith has its extremist bigots who use religion as nothing more than an excuse for their acts of evil.
I think this point can't be emphasized enough.
Are the guys that bomb abortion clinics fair representitives of Christianity? No, I don't think so.
If these terrorists were in a different country, they'd probably be a different religion. That doesn't matter. They use things like religion and nationalistic pride to give credibility to their actions.
To give a computer example, these terrorists are like the little punks that write viruses. They can say that they're raising awareness about some issue, but the fact of the matter is that they get off by causing destruction. Relatively minor destruction in the case of virus-writers, but destruction just the same.
Um, that's perfect topazes, not diamonds. Current reports suggest that a magic find percentage over 200 doesn't do much good, so there's no need to go overboard.
And they 'balanced' telekenesis so that you can only pick up minor items (like potions). This is very annoying in single player mode, where there is no one to steal drops from...
As for the economy aspect, I think it's kinda dumb to spend real money on Diablo II items. However, some people want uber-characters than can walk into any game and kill other players. Or maybe they're not good enough to take on Hell mode without help.
Dunno. Spending money on Diablo II items isn't any worse than spending it on an astrologer or the lottery.
At any rate, the economy associated with Diablo II has been interesting. There's been a lot of shifts in the marketplace in response to bugs allowing duplication of items, and the changes to gambling rules. Some moderately valuable items (like the Stone of Jordan ring or perfect skulls) became the new currency for a while. SoJs have become much more rare these days, and aren't used as currency as much. Pskulls are an interesting currency, because they are constantly being generated, but also constantly being used up.
Interestingly, gold (the currency inside the game) isn't often used for trading, because it isn't valuable enough!
I've also seen people auctioning off fully equipped high-level characters on eBay. However, with the balance changes in Lord of Destruction, some of these characters (which might have sold for hundreds of dollars) are now nearly worthless.
All in all, it's not too easy to base your economy on factors (like rarity) that can be changed at the whim of some programmers.
The key thing (with either corn or silver) is that there are end products in the process that have a consistent quality.
Anytime you're generating a large mass of something (like corn cobs) that has a consistent quality to them, there is potential to use it for something else.
If you take this potential raw material, and just throw it in the ground (or burn it), you're wasting the energy used to refine it. Sure, that energy was earmarked for refining the pricipal product (corn or silver), but as a result of that process you have also refined another potential product (corn cobs, or whatever is left of the silver ore). It usually just takes a little cleverness and effort to put this other raw material to good use too.
First a word of advice: If you're getting paid, and the company you're working for isn't about to go out of business, then strongly consider staying where you are.
Perhaps you can approach the problem a different way. You could try talking to your boss about the issues you're dealing with; it seems the worst one is repetitious nature of the code you're working on.
Anytime you're doing something repetitive with a computer it's usually boring, and it's a sign that you're doing something wrong.
Perhaps you can work with your company to develop a more abstract toolkit for your application area. If your programming lanugage/library doesn't support abstraction very well, perhaps you can come up with a code generator.
It may be easy for you to push out quick hacks. But how easy is it to write a program that can do the same? That could be a good challenge, and it would benefit the company because they could complete projects quicker. You might also get to use some new techniques or tools.
An employee who's constantly eliminating his own job is highly valued by good managers. Not that really good managers are all that common either...
The question is, how many different computer product lines can/should the combined company support?
I only count a few: Laptops (consumer and corporate), desktops (consumer and corporate), standard servers (small, medium, and large), and high-availability systems (large and gargantuan).
Both Compaq and HP currently have product lines for all these markets. Sometimes more than one. The combined company will have two or three product lines for each market. Can you smell trouble?
Even IBM wasn't able to pull this off forever. The combined company must figure out how to consolidate its product lines... especially because some of them are in direct competition with each other.
Take high-availability for example. Does it make sense to move all the Tandem customers to PA-RISC? Or the other way around? Or does it make sense to port all this stuff to Linux, which could eventually run the entire server line? Which strategy will be faster? Easier? Cheaper? Which will be less porting effort? What will the current customers tolerate?
Dunno. Tough call. I'm glad I'm not sitting in the big chair for this one.
...but would Linux even exist in anything approaching it's present magnificent state without the antagonism of Microsoft?
If you truly believe this, then you know little of Linux, and the hacker culture.
Linux (the kernel) exists because Linus thought it would be cool to write it. Other people also thought it was cool, and they decided to help out.
The first Linux distributions were originally created because hackers wanted them. Not because they hated M$, but because they wanted something good to work on, tinker with, and do development in.
Fast forward to today, and we have a lot of people, even suits, who are starting to recognize the benefits of Open Source / Free Software.
...will it [the community] be able to hold together?
Hah, like we're all that together now! Remember, cooperation is good and competition is good. We need both.
Ah, young grasshopper. A good interface design can take you a long way towards making a competing product.
It is often insightful to see what was put into the hardware interface, and what was left out. Often these can provide significant clues towards the design of a better product.
They're selling digital books over the Internet, at approximately $2.50 each, way less than the paperback version.
I've had a RocketReader since they first came out. However, I never purchased any books for the device. I didn't want to deal with the hassle of re-licensing books if my reader ever bit the dust (which is likely considering how many times I've dropped it already). Too much like Divx (the failed DVD format) for my tastes.
In contrast, Baen sells books unencrypted, in formats like RTF and HTML. I feel safe buying these books, because I know I'll always have software available to read the format. This is a very good thing.
There's a very good essays on their site explaining their philosophy, and why they think it works. Jim Baen clearly sees which way things are going because of advances in technology. He gets it. Which shouldn't be too suprising since he's a SF publisher...
TPM painted a much more interesting political behind-the-scenes picture, especially considering the larger context (Palpatine's rise to power).
That, in my opinion was the only thing (aside from some of the action sequences) that they got right in Episode I.
Palpatine is the pinnacle of lawful evil (to use a D&D concept). He is cold, ruthless, and very, very clever. In Ep1, he arranges events so that no matter what happens, he wins. Classic.
That's a very good point. I had never thought of it that way, but you're right. Han is supposed to be a lying, cheating, stealing, arrogant braggart. But he starts to see the bigger picture, and he learns what is really important.
Please don't do this. Transcoding almost always leads to very low quality files...
I can certainly attest to this. A friend converted some MP3s to put on his Sony minidisc player (which uses it's own compression scheme). The results sounded awful. It literally sounded as if the song had run over by a truck or something.
Well, I don't see the computational efficiency of humans (or future AIs) as being a problem.
It takes human-level intelligence to correlate interesting information together (design of proposed chemical plant, mapping of local water table). But it doesn't take human-level intelligence to actually run the numbers and discover that there's a problem (arsenic levels in drinking water over EPA guidelines).
Future AIs will be able to do the same things we do now. Except that the AI will be directly wired to unbelievably fast parallel supercomputers. (Dare I say Beowulf Cluster?)
These AIs will be able to simulate complex weather systems as easily as you can calculate a mortgage table in Gnumeric.
Slashdot Mirror
Mozilla is a hog, and if you're opening up a few documents, expect OpenOffice to use more too. I'd figure at least 100MB per user. Now we're talking 50GB of RAM.
I don't know what the author's been smoking, but I'd never put all the users on just one box. If a lab PC goes down, the user can switch to another. If your one and only Unix server goes down, everyone goes home. It would be better to split the processors between two smaller boxes, than to put everything in one.
The processors and memory are the largest part of the 4800's cost (especially in the configurations we're talking about. The chassis, backplane, and power supplies are relatively cheap.
And what the heck is going on with a SPARCstation 10 as the management console? Excuse me, but those have been discontinued for how many years? He mentions Office XP, so it's not as if this "report" was written in 1995. Sheesh.
I am a total Unix/Linux advocate, but this "report" is completely bogus.
Well, the guy was using a laptop, so that's not really an option, is it?
I've gone the build-it-yourself route. I selected the sound card, video, DVD, etc. Because it's a game system, I install Win98. Well, Diablo II runs like crap on it, other games are fine. I check the driver version, I check the message boards, I check Blizzard's site. No help.
As a lark, I install Windows NT 4.0. Diablo II now runs like a champ under that. Exact same hardware, same release of the drivers. But a big difference between 98 and NT 4.0. Why? who knows.
Bah. My next game machine will be a pre-built box . I don't want to deal with that crap any longer.
Heck, I may abandon Windoze altogether, and try to survive on Linux for gaming. Even then, I'll go with a pre-built box that has a proven configuration.
Two scenarios:
1. You get an e-mail virus written for your e-mail application. Your OS has multi-user security, so that the system binaries aren't affected. You log in as the root user, and clean up the affected files.
2. You get an e-mail virus. Since your OS doesn't have memory protection, it copies itself onto any executable it finds. You have to re-install your operating system because it's impractical to undo all the changes the virus made.
You pick.
On the surface, this appears to be a terrible situation.
If you read a little further into the case, it doesn't seem like miscarriage of justice.
SBCCI and the like are non-profit corporations that earn money by publishing building standard codes. Cities then buy copies of the codes, and incorporate them into law, which is much easier than creating huge volumes of building codes by oneself.
You're allowed to look at the building codes for free; you can go to the city office. You can even make copies there. However, you're not allowed to publish copies the building code.
Do a web search to find out more about this case.
At any rate, that situation doesn't apply to what we're discussing with the W3C. The building codes themselves aren't proprietary... anyone is allowed to build a building according to the code.
People from the W3C have "acknowledged" that the Internet's growth has been due to open standards.
This isn't not even half the story. The Internet would not exist without open standards promoted by bodies like the IETF and W3C (until now).
In the 1980's and early 1990's there were a number of network protocols in use: DECNet, VINES, NETBIOS/NETBEUI (shudder), IPX/SPX, SNA, and more. None of them initially would have been as scalable as TCP/IP, however, if any of them had been truly open, it might have been possible to fix them.
But none of those other protocols were open... and where are they now? Nowhere.
It's the same situation for hypertext protocols. People and companies have proposed substantial improvments onto existing protocols. A notable example of this is Hyper-G, which was then commercialized by Hyperwave.com. It fixes a lot of problems with navigation, and stuff like broken links. However, there was never a free and open implementation, and so it has languished in obscurity for the last 5 years.
My question to the W3C is this: Do they have any evidence that proprietary protocols will foster continued growth of the Internet and the applications that run on top of it?
I'm sure someone's already come up with a script that can automatically check the GPG signatures of downloaded packages.
Of course, I would never run automatic software updates on a production server. That might be fine for desktop machines though.
This is indeed worrisome, but some proliferation of licenses is necessary.
For example:
I was "around" (reading the mailing lists) when the Eiffel Forum License was created. Because of the nature of OO software development, it was felt that existing licenses (such as the LGPL) would not be appropriate. I don't recall all the discussions, but there were some good reasons why people didn't want to use the existing ones.
This is the "downside" to freedom; everyone tends to go off in their own direction. Of course M$ and the other industry players would be glad to consolidate end-user licenses down to just one. That would make things very simple... but not desirable or pleasant.
Excluding one time expenses is GAAP.
That is very true.
However, being in business myself, I know that a loss is a loss is a loss. It doesn't matter what you call it, it doesn't change the fact that there's more money going out the door than coming in.
Just because things look good on paper doesn't really mean that much. What really counts is how much money you have in the bank.
So, at any rate, it's good that I swiched to Debian last year. :-)
The trouble with stack machines is that they have a worse Von Neumann bottleneck than register machines. Everything has to go through the top of the stack.
Hmmmm.... I wonder about that. If you designed a CPU for Forth (or another stack-based language), you could have something where the first, say, 10 elements of the stack are in registers, and the rest in memory. As the stack grows, the bottom elements are moved off into main memory. Vice versa for shrinking. If done cleverly, there wouldn't be much need for copying data around.
Of course, I'm sure someone's already thought of this...
Well, you won't catch me spending a lot of time in front of a microwave transceiver.
Microwave ovens are optimized to heat water, but other types of microwaves (used in communications) can still do that too. Plus, I don't know what effect such high power levels would have (besides heating) but there might be some.
Good enough reason to be careful!
The only problem with those "special guns that won't puncture steel" is that aircraft are mostly made out of aluminum these days. Better off with tazers, beanbags and such. Pepper spray would suck in such an enclosed environment.
As for hijackers intentionally crashing the plane they've taken over, it is perhaps not totally without precedent.
There was a hijacking of an aircraft out of India a few years ago. The guys were idiots, and forced the pilot to keep flying until the plane ran out of fuel.
Fortunately, the plane had just crossed over the shore, and the pilot tried to ditch in the ocean. Unfortunately, one of the hijackers grapped the flight yoke at the last second, and prevented the pilot from making a smooth water landing.
It probably didn't make as much news because only 1/3 of the people died.
At any rate, it won't as easy to do what they did again. I know that if I see someone trying to take over the plane I'm in, I'll do my best to immediately stop it. I'm not much of a fighter, but I'd rather take my chances with being shot a point blank range than having the plane I'm in crash into a building. At least I'll use up a bullet, and maybe give the guy behind me a chance...
The world got a lot colder and meaner Tuesday.
Every faith has its extremist bigots who use religion as nothing more than an excuse for their acts of evil.
I think this point can't be emphasized enough.
Are the guys that bomb abortion clinics fair representitives of Christianity? No, I don't think so.
If these terrorists were in a different country, they'd probably be a different religion. That doesn't matter. They use things like religion and nationalistic pride to give credibility to their actions.
To give a computer example, these terrorists are like the little punks that write viruses. They can say that they're raising awareness about some issue, but the fact of the matter is that they get off by causing destruction. Relatively minor destruction in the case of virus-writers, but destruction just the same.
Um, that's perfect topazes, not diamonds. Current reports suggest that a magic find percentage over 200 doesn't do much good, so there's no need to go overboard.
And they 'balanced' telekenesis so that you can only pick up minor items (like potions). This is very annoying in single player mode, where there is no one to steal drops from...
As for the economy aspect, I think it's kinda dumb to spend real money on Diablo II items. However, some people want uber-characters than can walk into any game and kill other players. Or maybe they're not good enough to take on Hell mode without help.
Dunno. Spending money on Diablo II items isn't any worse than spending it on an astrologer or the lottery.
At any rate, the economy associated with Diablo II has been interesting. There's been a lot of shifts in the marketplace in response to bugs allowing duplication of items, and the changes to gambling rules. Some moderately valuable items (like the Stone of Jordan ring or perfect skulls) became the new currency for a while. SoJs have become much more rare these days, and aren't used as currency as much. Pskulls are an interesting currency, because they are constantly being generated, but also constantly being used up.
Interestingly, gold (the currency inside the game) isn't often used for trading, because it isn't valuable enough!
I've also seen people auctioning off fully equipped high-level characters on eBay. However, with the balance changes in Lord of Destruction, some of these characters (which might have sold for hundreds of dollars) are now nearly worthless.
All in all, it's not too easy to base your economy on factors (like rarity) that can be changed at the whim of some programmers.
The key thing (with either corn or silver) is that there are end products in the process that have a consistent quality.
Anytime you're generating a large mass of something (like corn cobs) that has a consistent quality to them, there is potential to use it for something else.
If you take this potential raw material, and just throw it in the ground (or burn it), you're wasting the energy used to refine it. Sure, that energy was earmarked for refining the pricipal product (corn or silver), but as a result of that process you have also refined another potential product (corn cobs, or whatever is left of the silver ore). It usually just takes a little cleverness and effort to put this other raw material to good use too.
First a word of advice: If you're getting paid, and the company you're working for isn't about to go out of business, then strongly consider staying where you are.
Perhaps you can approach the problem a different way. You could try talking to your boss about the issues you're dealing with; it seems the worst one is repetitious nature of the code you're working on.
Anytime you're doing something repetitive with a computer it's usually boring, and it's a sign that you're doing something wrong.
Perhaps you can work with your company to develop a more abstract toolkit for your application area. If your programming lanugage/library doesn't support abstraction very well, perhaps you can come up with a code generator.
It may be easy for you to push out quick hacks. But how easy is it to write a program that can do the same? That could be a good challenge, and it would benefit the company because they could complete projects quicker. You might also get to use some new techniques or tools.
An employee who's constantly eliminating his own job is highly valued by good managers. Not that really good managers are all that common either...
The question is, how many different computer product lines can/should the combined company support?
I only count a few: Laptops (consumer and corporate), desktops (consumer and corporate), standard servers (small, medium, and large), and high-availability systems (large and gargantuan).
Both Compaq and HP currently have product lines for all these markets. Sometimes more than one. The combined company will have two or three product lines for each market. Can you smell trouble?
Even IBM wasn't able to pull this off forever. The combined company must figure out how to consolidate its product lines... especially because some of them are in direct competition with each other.
Take high-availability for example. Does it make sense to move all the Tandem customers to PA-RISC? Or the other way around? Or does it make sense to port all this stuff to Linux, which could eventually run the entire server line? Which strategy will be faster? Easier? Cheaper? Which will be less porting effort? What will the current customers tolerate?
Dunno. Tough call. I'm glad I'm not sitting in the big chair for this one.
If you truly believe this, then you know little of Linux, and the hacker culture.
Linux (the kernel) exists because Linus thought it would be cool to write it. Other people also thought it was cool, and they decided to help out.
The first Linux distributions were originally created because hackers wanted them. Not because they hated M$, but because they wanted something good to work on, tinker with, and do development in.
Fast forward to today, and we have a lot of people, even suits, who are starting to recognize the benefits of Open Source / Free Software.
Hah, like we're all that together now! Remember, cooperation is good and competition is good. We need both.
Ah, young grasshopper. A good interface design can take you a long way towards making a competing product.
It is often insightful to see what was put into the hardware interface, and what was left out. Often these can provide significant clues towards the design of a better product.
Yes. Go to http://www.baen.com.
They're selling digital books over the Internet, at approximately $2.50 each, way less than the paperback version.
I've had a RocketReader since they first came out. However, I never purchased any books for the device. I didn't want to deal with the hassle of re-licensing books if my reader ever bit the dust (which is likely considering how many times I've dropped it already). Too much like Divx (the failed DVD format) for my tastes.
In contrast, Baen sells books unencrypted, in formats like RTF and HTML. I feel safe buying these books, because I know I'll always have software available to read the format. This is a very good thing.
There's a very good essays on their site explaining their philosophy, and why they think it works. Jim Baen clearly sees which way things are going because of advances in technology. He gets it. Which shouldn't be too suprising since he's a SF publisher...
I like 'ogg' because it reminds me of netrek playing. To 'ogg' is to blow up an enemy ship that has armies. Great fun.
TPM painted a much more interesting political behind-the-scenes picture, especially considering the larger context (Palpatine's rise to power).
That, in my opinion was the only thing (aside from some of the action sequences) that they got right in Episode I.
Palpatine is the pinnacle of lawful evil (to use a D&D concept). He is cold, ruthless, and very, very clever. In Ep1, he arranges events so that no matter what happens, he wins. Classic.
That's a very good point. I had never thought of it that way, but you're right. Han is supposed to be a lying, cheating, stealing, arrogant braggart. But he starts to see the bigger picture, and he learns what is really important.
Please don't do this. Transcoding almost always leads to very low quality files...
I can certainly attest to this. A friend converted some MP3s to put on his Sony minidisc player (which uses it's own compression scheme). The results sounded awful. It literally sounded as if the song had run over by a truck or something.
Well, I don't see the computational efficiency of humans (or future AIs) as being a problem.
It takes human-level intelligence to correlate interesting information together (design of proposed chemical plant, mapping of local water table). But it doesn't take human-level intelligence to actually run the numbers and discover that there's a problem (arsenic levels in drinking water over EPA guidelines).
Future AIs will be able to do the same things we do now. Except that the AI will be directly wired to unbelievably fast parallel supercomputers. (Dare I say Beowulf Cluster?)
These AIs will be able to simulate complex weather systems as easily as you can calculate a mortgage table in Gnumeric.
I was under the impression that most of these machines are standalone, and not clustered together.