Hey, FWIW as an aspiring author I rather like the idea that I can develop something that no one can take away from me.
Indeed, while everyone is belly acheing about how long copyrights last, the only ones who really benefit from a copyright expiring are those who mint cheap copies. Sure it sucks that the Girl Scouts need an ASCAP license to sing 'God Bless America.' But there is nothing keeping some counselor or scout from writing another song.
Let's face it, we have gotten so innured about buying pre-made books, and pre-made music that we have almost forgotten that there was a time when people would just make this stuff up. (Sigh, I go to a folk festival every year and it seems that most wandering minstrals only know Beatles tunes.)
Data is a jumble of facts. Information is the stuff that ties all those facts together into a useful representation of the world.
Those who simply collect and warehouse data have nothing. It's like the folks who fill their house with useless junk, and then start dumping garbage on their lawn.
Now, compare that to a museum. They have a lot of what would be otherwise useless junk. But they track where it came from. They track who owned it. They track what part the thingy played in history. They also package it, preserve it, and display it in a way that can be found again.
And I work at one. Museums throw things out all the time. It is the process of filtering, filing, and preservation that turns data into information. Much the same way a brick house is not the same as a pile of dirt, despite the fact that chemically they are the same.
What the MCSE courses for the XBox? Egads, I guess Sony is going to try to top that with tech school certs on the PS/2, and Nintendo will try to follow along 3 years later with an associates degree from 'Mario University.'
Hey, my manager at the pizza place was just a damn thief. The good ones don't bother snorting coke, they just sink it right into cars, boats, and houses.
Ooo, even better. Patent the process of patenting the method for paying open source workers on the web. That way, when someone finally comes along with a working model, we can sue them for taking our idea. Of course, without having to work out all the nasty details ourselves.
Remember IBM was Microsoft before there was microsoft. These are the guys who invented FUD as we know it. Being a monopolist seems to bring out the worst in people. It's like the damn ring of power.
I wouldn't sit up nights worrying about the process for one simple reason. Most OSS projects are run ostensibly by volunteers. Sure they have a day job, but it's still their copyrights and names in the credits. (Companies insist on copyrighting in their own name. They also tend to filter out individual contributorship.)
For the record chip design WAS an area of my expertice, and I have had PhD's confirm my suspicions that Intel's new stuff was bogus. Indeed, I've done the simulations myself.
Come on, this is the same company that is pitching 'centrino' as a new platform. It's nothing more than integrating 802.11 wireless into the chipset.
This is the same company that produced the 386SX. It was a 16bit chip with a 32 bit bus. They later got it right with the 386 DX, and then continued to sell the crippled chip. They had such a success with the SX/DX thing they tried it with the 486. The 486SX was a DX chip that they etched out the FPU unit out of. They you would later buy the '487' co-processor which was simply a 486 chip that deactivated the original chip and took over.
Oh yea, and let's not forget the whole Celeron cache size debacle, nor the fact the PII and PIII are essentally the same. The Xeon is a PIII with extra cache. Ooooooo.
The only reason Intel needs a predictor is because the chip moves so much faster than the RAM. The machine spends more time guessing what it should do than actually doing it. Why? To dance around the fact the most people would never see the difference in performance between an 800Mhz machine and 4Ghz machine.
It is all about seperating you from your almighty dollar, friend.
MPEG is a memory bound task. It's also a disk bound task. I can't think of many computers that can keep gigabytes of data in RAM, let alone in the registers of the processor. The hardest part about playing DVD's was getting the information across the IDE bus in real time. Ok, and before we had 266Mhz memory busses, getting the data through to the processor in real time.
SQL is slaved to the network card. It too is disk and memory dependent. Sure it will try to cache as much of the stuff stored on disk in memory, but RAM is finite and fleeting. Plus it has to store the data somewhere.
Now SQL can scale to multiple processors because you are almost never running only one query at a time. If 2 or more queries are running, each processor can do one in parallel.
As far as "optimizing" your c++ code, don't bother. Yes, there are inefficient algorythems out there that should be avoided. In general, the compiler takes care of the optimizations for you. The compiler also takes care of grouping code together, and the OS controls when it's swapped in and out of the processor. About the only thing I've seen come out of developers writing "optimal" code is undecipherable code.
And no, it didn't really run any better.
You have to have an intimate understanding of your processor architecture before you can make decisions about quanta. And then, you are generally coding in assember.
You have to remember that a garden variety PC is a very unpredictable environment. You have network packets coming in, mouse events, keyboard presses, USB chatter, DMA access, every event generates and interrupt that requires the processor to stop what it's doing, and start the pipeline over again.
Yes, it will be very useful for calculating seti@home and missile simulations, namely because the task it absolutely predictable and requires very little disk or network I/O. Everything else slogs it down. It's like a Hummer in Tokyo. Sure it can go off-road, but where are you going to park it?
There are limits to how much you can route around the brain-deadedness of a design.
Having to deal with a 30+ stage pipeline makes the compiler VERY complicated, and it also magnifies any errata.
Another thing to consider is that the processor is continually interupted by I/O tasks on a desktop PC. It has to drop what it's doing, respond to the interrupt, and then re-load the state from where it was at. Simple things like disk access and network packets generate interrupts.
A P4 would work great in a supercomputer. And that's about the only application it would be good for. Well, decent at. A PPC would eat it for lunch because it has a few extra ALU's and FPU's.
On the contrary. He has not opened himself (or his company) to anything. This is no different than someone trying to invoice me for a service that was never provided. There is no legal relationship between the 2 parties. There is no pending litigation.
The day I lose the right to call shit as shit is the day I move somewhere else.
Now as a network admin, I am in a position of trust. I can more or less poke around the system at will, read any files I'd like, and sift through everyone's email. While it is techically possible, if I were ever caught doing this I would be fired.
I'm not even sure I would get to clean out my desk.
This is not a matter of Joe Hacker forwards an internal memo. This is a matter of one competing faction within an organization abusing his or her access to a computer system. That is bad enough. They had to take it a step further and PUBLICIZE the information they found.
Joe Hacker is an outsider acting on his own. The Halloween memos and such, he has an informant on the inside. He may embarrass a company. He may steer a lawsuit. The worst damages are monetary.
Jane Insider, on the other hand, is committing betrayal. She is seeking to influence elections and the operations of government. All this while working for an elected official.
Both Joe and Jane should probably get an extended stay at Uncle Sam's Federal Resort. Joe for theft, Jane for treason. It doesn't matter WHAT party you are working for. You do not fold mutilate or spindle and elected official's documents.
I hate to tell you this, but the Uranium was radioactive before we refined it. It will still be radioactive if we leave it in the ground. The joy of nuclear power is that it is pent up energy that is out there for the taking.
Besides, anything with a half life THAT long isn't very radioactive, is it? Think about it, radiation is the rate at which something emits neutrons. In the process of emitting neutrons, the material decays into another, more stable form. Highly radioactive materials degrade very fast. Some elements we have created artificially last a fraction of a second.
I co-oped in a Steel mill. I had fun, but looking bakc it was f@$!ing scary. A lot of fun at the time, but not something I'd do today with wife and kid waiting at home for me.
You had to watch for forklifts whizzing by at 40mph carrying a few thousand pounds of material. Most of the systems I was working on were sealed in large metal cabinets. Despite being behind a concrete barrier, one of the ones I used to change tapes out on had a large dent up top...
My strangest day was going out and retreiving a serial number off an old machine. I had to go to the melt shop. I was greated by a large sign: IF YOU DO NOT WORK IN THIS AREA, STAY AWAY.
Well, I had clearence. Sort of. My boss told me to go. I walked across a floor caked with chemical residue, and over to a 50 foot tower in the middle of a huge hanger like building. From the top of the tower poured a river of liquid steel. At the first landing was a sign: DANGER: RADIATION.
No, not nuclear radiation. Heat radiation. If you stand there long enough, the heat coming off the steel as it goes by is enough to burn you.
I made my way to the top of this tower, and opened the door to the control room. In it were 10 or so people having a luncheon of roast turkey and ham.
The system I needed to access was sealed under a counter that I had to disassemble with my pocket knife. I did get a free lunch out of the deal though.
Alright, I enjoyed it. Who can honestly say anything bad about watching a volkswagon sized block of steel being drop forged into a cylinder 40 feet long by a huge device that looks like something from a horror movie. The arc furnaces were fun to watch too. They had to use LCD's there because the magnetic fields put out by the device completely scrambled CRT signals.
I worked there over the winter. Most of the building were hotter than the 8th ring of hell. Outside was as cold as a witches tit. Anywhere in between was the bastion of the forklifts.
Slashdot Mirror
Amazing Grace
How sweet the sound
that save a wretch like me
I once was lost
but now I'm found
was blind but now I see.
Indeed, while everyone is belly acheing about how long copyrights last, the only ones who really benefit from a copyright expiring are those who mint cheap copies. Sure it sucks that the Girl Scouts need an ASCAP license to sing 'God Bless America.' But there is nothing keeping some counselor or scout from writing another song.
Let's face it, we have gotten so innured about buying pre-made books, and pre-made music that we have almost forgotten that there was a time when people would just make this stuff up. (Sigh, I go to a folk festival every year and it seems that most wandering minstrals only know Beatles tunes.)
Actually write a letter to your elected "politician" and tell me that again with a straight face.
Data is a jumble of facts. Information is the stuff that ties all those facts together into a useful representation of the world.
Those who simply collect and warehouse data have nothing. It's like the folks who fill their house with useless junk, and then start dumping garbage on their lawn.
Now, compare that to a museum. They have a lot of what would be otherwise useless junk. But they track where it came from. They track who owned it. They track what part the thingy played in history. They also package it, preserve it, and display it in a way that can be found again.
And I work at one. Museums throw things out all the time. It is the process of filtering, filing, and preservation that turns data into information. Much the same way a brick house is not the same as a pile of dirt, despite the fact that chemically they are the same.
What the MCSE courses for the XBox? Egads, I guess Sony is going to try to top that with tech school certs on the PS/2, and Nintendo will try to follow along 3 years later with an associates degree from 'Mario University.'
Hey, my manager at the pizza place was just a damn thief. The good ones don't bother snorting coke, they just sink it right into cars, boats, and houses.
Ooo, even better. Patent the process of patenting the method for paying open source workers on the web. That way, when someone finally comes along with a working model, we can sue them for taking our idea. Of course, without having to work out all the nasty details ourselves.
Take Linus. He works for OSDL. They keep him in a lifestyle to which he has been accustomed in exchange for continuing to work on Linux.
Man, how much of the kernel do you have to write to get a deal like that?
I wouldn't sit up nights worrying about the process for one simple reason. Most OSS projects are run ostensibly by volunteers. Sure they have a day job, but it's still their copyrights and names in the credits. (Companies insist on copyrighting in their own name. They also tend to filter out individual contributorship.)
Remember, H.A.L. is I.B.M. decremented on letter each. Kubric and Clarke swear it was a coincindence...
Come on, this is the same company that is pitching 'centrino' as a new platform. It's nothing more than integrating 802.11 wireless into the chipset.
This is the same company that produced the 386SX. It was a 16bit chip with a 32 bit bus. They later got it right with the 386 DX, and then continued to sell the crippled chip. They had such a success with the SX/DX thing they tried it with the 486. The 486SX was a DX chip that they etched out the FPU unit out of. They you would later buy the '487' co-processor which was simply a 486 chip that deactivated the original chip and took over.
Oh yea, and let's not forget the whole Celeron cache size debacle, nor the fact the PII and PIII are essentally the same. The Xeon is a PIII with extra cache. Ooooooo.
The only reason Intel needs a predictor is because the chip moves so much faster than the RAM. The machine spends more time guessing what it should do than actually doing it. Why? To dance around the fact the most people would never see the difference in performance between an 800Mhz machine and 4Ghz machine.
It is all about seperating you from your almighty dollar, friend.
MPEG is a memory bound task. It's also a disk bound task. I can't think of many computers that can keep gigabytes of data in RAM, let alone in the registers of the processor. The hardest part about playing DVD's was getting the information across the IDE bus in real time. Ok, and before we had 266Mhz memory busses, getting the data through to the processor in real time.
SQL is slaved to the network card. It too is disk and memory dependent. Sure it will try to cache as much of the stuff stored on disk in memory, but RAM is finite and fleeting. Plus it has to store the data somewhere.
Now SQL can scale to multiple processors because you are almost never running only one query at a time. If 2 or more queries are running, each processor can do one in parallel.
As far as "optimizing" your c++ code, don't bother. Yes, there are inefficient algorythems out there that should be avoided. In general, the compiler takes care of the optimizations for you. The compiler also takes care of grouping code together, and the OS controls when it's swapped in and out of the processor. About the only thing I've seen come out of developers writing "optimal" code is undecipherable code.
And no, it didn't really run any better.
You have to have an intimate understanding of your processor architecture before you can make decisions about quanta. And then, you are generally coding in assember.
Yes, it will be very useful for calculating seti@home and missile simulations, namely because the task it absolutely predictable and requires very little disk or network I/O. Everything else slogs it down. It's like a Hummer in Tokyo. Sure it can go off-road, but where are you going to park it?
Having to deal with a 30+ stage pipeline makes the compiler VERY complicated, and it also magnifies any errata.
Another thing to consider is that the processor is continually interupted by I/O tasks on a desktop PC. It has to drop what it's doing, respond to the interrupt, and then re-load the state from where it was at. Simple things like disk access and network packets generate interrupts.
A P4 would work great in a supercomputer. And that's about the only application it would be good for. Well, decent at. A PPC would eat it for lunch because it has a few extra ALU's and FPU's.
The day I lose the right to call shit as shit is the day I move somewhere else.
Now as a network admin, I am in a position of trust. I can more or less poke around the system at will, read any files I'd like, and sift through everyone's email. While it is techically possible, if I were ever caught doing this I would be fired.
I'm not even sure I would get to clean out my desk.
This is not a matter of Joe Hacker forwards an internal memo. This is a matter of one competing faction within an organization abusing his or her access to a computer system. That is bad enough. They had to take it a step further and PUBLICIZE the information they found.
Joe Hacker is an outsider acting on his own. The Halloween memos and such, he has an informant on the inside. He may embarrass a company. He may steer a lawsuit. The worst damages are monetary.
Jane Insider, on the other hand, is committing betrayal. She is seeking to influence elections and the operations of government. All this while working for an elected official.
Both Joe and Jane should probably get an extended stay at Uncle Sam's Federal Resort. Joe for theft, Jane for treason. It doesn't matter WHAT party you are working for. You do not fold mutilate or spindle and elected official's documents.
W@erg@e
Much better to send it to an Engineering school. Or better yet, a bunch of Engineering school dropouts.
Besides, anything with a half life THAT long isn't very radioactive, is it? Think about it, radiation is the rate at which something emits neutrons. In the process of emitting neutrons, the material decays into another, more stable form. Highly radioactive materials degrade very fast. Some elements we have created artificially last a fraction of a second.
I just don't think the punchline was particularly funny.
Any way that Dubya can pay Steve a visit first?
Dred Lord.
You had to watch for forklifts whizzing by at 40mph carrying a few thousand pounds of material. Most of the systems I was working on were sealed in large metal cabinets. Despite being behind a concrete barrier, one of the ones I used to change tapes out on had a large dent up top...
My strangest day was going out and retreiving a serial number off an old machine. I had to go to the melt shop. I was greated by a large sign: IF YOU DO NOT WORK IN THIS AREA, STAY AWAY.
Well, I had clearence. Sort of. My boss told me to go. I walked across a floor caked with chemical residue, and over to a 50 foot tower in the middle of a huge hanger like building. From the top of the tower poured a river of liquid steel. At the first landing was a sign: DANGER: RADIATION.
No, not nuclear radiation. Heat radiation. If you stand there long enough, the heat coming off the steel as it goes by is enough to burn you.
I made my way to the top of this tower, and opened the door to the control room. In it were 10 or so people having a luncheon of roast turkey and ham.
The system I needed to access was sealed under a counter that I had to disassemble with my pocket knife. I did get a free lunch out of the deal though.
Alright, I enjoyed it. Who can honestly say anything bad about watching a volkswagon sized block of steel being drop forged into a cylinder 40 feet long by a huge device that looks like something from a horror movie. The arc furnaces were fun to watch too. They had to use LCD's there because the magnetic fields put out by the device completely scrambled CRT signals.
I worked there over the winter. Most of the building were hotter than the 8th ring of hell. Outside was as cold as a witches tit. Anywhere in between was the bastion of the forklifts.
I'm waiting for them to retroactively sue AT&T for licensing out Unix to begin with.
Hey, I have some good friends that are Jehova's Witness's. Equating them with McBride is like equating motherhood with prostitution.