Instead we're turning into a rabble of peasants and share-croppers slaving for, and kowtowing to, the modern day Lords of Corporatism. And we put on our chains so willingly!
The really sickening part is who owns all those corporations. It's the very same rabble of peasants who's being ground under foot.
The working class has far more total wealth than the upper class just because there's so many more working class people. That money is mostly held as corporate shares, either through CDs at the bank (the bank re-invests that money) or mutual funds.
It is the average person enslaving themselves here.
I still say the only way you had a 60% failure rate out of the box is if someone dropped the box. My HDs have almost always been out of those cases from tiny stores that buy bulk cases of parts intended for OEM computer assembly and sell them at retail. The only ones I've bought in retail packaging are when big box stores have mail in rebate specials that make them even cheaper.
Stop buying off eBay. I've bought about 10 Maxtor and Seagate IDE HDs for personal use in the past 5 years, and I've had 0 failures. I know that's pretty lucky (and I do keep carefuly backups), but come on, 60% failure right out of the box? Who dropped the box before you tried them?
Of course now that I've said that, 60% of my drives will probably die in the next few days:).
No, my point all along is that it is essential for any good programmer to know how to program in assembly language, simply because it's necessary to really understand how the underlying hardware works so you can write effective code.
Your blind statement that your inefficient coding of what you're saying is an efficient algorithm is better than any possible efficient coding of a less efficient algorithm simply proves that point.
If you had a strong foundation in assembly, you wouldn't have made such a broad statement, because you'd understand the large demands quicksort has for system resources. Recursion is computationally very expensive to do, but in your high level ivory tower, that's all hidden from you. By choosing a platform that is incapable of the requried recursion, I just highlighted the point.
I'm also not sure what you're calling an analogy.
I never said a PIC controller was the ultimate computing platform. In fact, I picked a specific task the PIC was essentially useless for. It also is pretty useless at bit banging gigabit ethernet or USB. The baseline units don't even have an integer multiplication instruction.
I'll take that bet...but since you choose the algorithms, I choose the architectures, and I choose a base-line PIC microcontroller. It has a 2-level deep hardware stack, Let's see your recursive javascript code run on that.
I use assembly every day to program embedded controllers. There are compilers for them, but I often use controllers with a few hundred instruction words of storage space and a few dozen bytes of ram. Compilers just aren't efficient enough.
At one time (and probably to this day) the US DOD specs used to require a certain number of passes of 0 and 1 bits followed by the writing of a specific bit pattern before a hard drive was considered to have been properly erased.
I find it hard to believe the US DoD is this lax on security. I used to work for the Canadian government, and we had to hammer a nail through the drive a certain number of times "according to the specs" to consider it properly erased.
Based on a binary floating-point number and switching system, it had all the attributes of today's computers, such as a control block, a memory, and a calculator. But it didn't have the ability to store the program in the memory together with the data because the memory was too small.
Modern computers don't necessarily have the program memory in the same space as the data memory. Machines using the Von Neumann architechture, such as a PC have a shared memory space. The newer Harvard architechture has separate program and data spaces.
There are many advantages to separating them. The main one is that you can concurrently fetch the data for the current instruction and the next instruction word from the two separate memory busses, effectively doubling your throughput. Also, you're going to want your memory to be a multiple of 8 bits wide, but there is no reason your instruction word should have this restriction.
Re:R/C car modding
on
Old Toy Modding?
·
· Score: 1, Interesting
He can't post pictures because's a BS post.
The is no mechanical throttle in an electric RC car to "tweak and superglue". It uses a PWM signal to an H-Bridge driver chip to the motor; all electronic connections. On a toy car, this is all done on one ASIC that gets RF in from the remote and outputs to the motor directly.
Toys are a great source of mechanical parts
on
Old Toy Modding?
·
· Score: 1
I've turned remote control cars in to various forms of robots; line following, sumo, etc.
It costs a fraction of the price of buying comparable parts from a normal supplier, and a lot of the mechanical work is done for you.
Jason ProfQuotes
If employees are willing to put in the hours, the company doesn't value the time anyway. If they cared about their employees, they wouldn't put such demands on them.
When I worked for a start-up, I was willing to put in extra hours as needed, but it was generally only needed to compensate for the gross mis-management of the company.
For example, we were developing a set-top video device, and there was only 1 test-model for the whole company. At one point, I needed to test some code on a wednesday morning, and my boss literally had me sit and watch for a chance to test it until friday evening. I wanted to do other work, but he explicitly said I was supposed to sit there and wait. On friday afternoon, he "authorized" me to come in on the weekend to do it, and acted like he was doing me a huge favour by letting me go in for no extra pay. Of course, I refused, but it also meant I was first on the chopping block when the company downsized a few months later.
When I was there, a typical work week was 70-80 hours (these people could have had a higher hourly wage at McDonalds), after the downsize, I kept in touch with some people there and it was closer to 100. 100 hours a week at a $50k/year (canadian) job; it's insane. It comes out to $10/hour for an exhausting an emotionally destructive job. Obviously these people have no life at all; the only people bringing dates/spouses to the christmas party was senior management.
I hated that so much I cancelled my satellite TV service. With the money, I signed up with netflix.
So far it's great. Most of what I borrow are TV series DVDs, but I spend a lot less time watching TV and yet I enjoy it a lot more since I'm watching exactly what I want instead of just staring at the TV with "nothing" on.
As the AC said, I negotiate the best possible price before we even start talking extras. The dealers walk right into that because the purpose of the extras is to artifically raise the price so you're left feeling like you got a better deal than you really did. They don't want to mention the extras until you've agreed on a base price.
The "transportation fee" only applies if you're stupid enough to pay it. It's one of the easiest to get waived. The other is "administrative fee" which is basically their advertising budge. Getting those waived is as simple as saying you're not paying them and getting up to leave.
If Deutsch's books were written in the same style as that article, I don't think it's fair to credit Deutsch. The article is almost impossible to understand, and I'm very familiar with the experiment, it's written in such a condescending tone and with a high noise to signal ratio, it's difficult to distill out what it's trying to say.
Also, it pretends to be a home parallel universe test in the introduction, but in the meat of the article, it just dismisses the key point with no, this can't be done in your dining room, so it's hardly a home test at all.
Basically, this article takes an elegant concept in quantum physics, and presents it in an needlessly complex way without even truly explaining what is going on. It also pretends to be explaining how to do the key part of the experiment at home and then glosses over the fact that it's skipping that.
I'm surprised poking holes through paper with a pin even works, even for a pin hole camera, people use aluminum foil for the hole so it can be smaller and more precise.
Using a hole though isn't even the best way to do the experiment. It will cause difraction to occur in 2 dimensions, it is a lot easier to observe what's going on if it happens in 1 dimension. In high school, we blackened a microscope slide with smoke from a candle, then made 2 parallel scratches by taping 2 flat razor blades together and drawing them across the blackened glass. It gave nice perfectly paralell slits a known distance apart (the thickness of the blade).
Again this just demonstrated the first part, not the part Deutsch claims proves parallel universes exist.
This is a very old experiment, and a well-known phenomenon. It was even one of the answers on slashdot's poll for favourite physics experiment (and my personal favourite).
Even the idea that it is proof of parallel universes is not original. Michael Crichton made that claim in his book Timeline. It's an excellent book (despite the horrible movie loosely based on it), but it is fiction.
I have had many a customer tell me that NO ONE has ever opened their computer (including them) - I open it up and there are screws missing, the magnetic sheild has fingerprints on it, etc etc.
Interesting, because I bought my Wallstreet brand new, it never had been opened by anyone (including me), until I opened it a few months ago to replace the sound board. There were no missing screws, but there were fingerprints on the shield over the CPU.
I have enough pc and notebook repair experience that I don't allow "service providers" to touch my computer; they're primarily concerned with speed, not quality. I've seen machines come back from "authorized" service centers with stripped or missing screws and even broken plastic parts; the hinge cluth and upper case plastic on the Wallstreet, are a pain to take off for example, it's quite common for things like that to get broken at the service center.
Of course the service tech doesn't say anything, leaves the broken part in, and even charges full price. I hate taking my computers in for repair under warranty, I would prefer to just get the replacement part and install it myself, but of course they won't allow that.
If I've added an upgrade to my notebook, why should I lose the warranty when an unrelated part fails due to a manufacturing or design defect within the warranty period?
The only difficulty I had was keeping track of all the screws.
I took apart my wallstreet to fix the faulty power connector on the sound card (it's a common defect, apple's design flaw). My solution to keeping track of the screws was to use my digital camera and take lots of pictures. Usually I'd place the screws beside the laptop in the same relative positions they belong in.
Also, I found This site very useful as a disassembly guide. The site posted is hardly a great new idea.
If you did anything with those drives (load a program, just leave it powered of, etc), the heads would fall out of alignment. It was fairly simple regular maintainence to realign the drive. The motor even had that sticker for the stobe freeze effect under 60Hz fluorescent lighting.
Any linux expert would have the network card working inless than 5 minutes in any of the development oriented distros.
Exactly my point. The OP said this is the year of linux, I said it isn't. It seems that you agree with me.
Your reply is kind of funny, you are blindly defending linux and attacking me for saying it's not the greatest thing ever invented, yet at the same time your attacks prove me point. If linux is not for me (with a CS degree), it sure isn't for most users.
I switched from Linux to Windows about 3 years ago. It was just a pain in the ass to try and do anything in linux. You want to install a network card? Better set aside a weekend to recompile the kernel and then spend countless frustrating hours trying to configure everything.
Plug the same card into a windows machine and you're browsing the web in under 5 minutes.
Just this week I'm trying to give linux another shot. I have an old powerbook I want to set up as a wireless web browsing station, and WiFi cards that it supports are all 5 year old tech which is still a lot more expensive than a new card, so using YellowDog linux seemed like an ideal solution. The main install went well enough, expect it's only got a 2 gig drive and the distro puts a lot of useless junk on (like PPoE) so there's no room for X.
Then I have to spend hours browsing poorly written howtos and websites to find a driver that *might* support my network card. I downloaded the prism2 sourcefiles (since it's running on a PPC, I couldn't find a binary version), and then when I try to 'make configure' it wants the kernel source code...okay, I download the kernel source code, and prism still says it's missing or incomplete. I even tried recompiling that source and I get the same problem.
It's just not worth it. I'll spend the $1600 on a new powerbook (that will run OSX and includes an apple WiFi card). Anything is better that continuing to struggle with linux.
This is not the year of linux yet. Maybe I'll try it again on a desktop in a couple of years.
It works well enough on the server; I do have a linux server that started as slackware 3.3, and maintaining server features is reasonably easy, but linux is still useless on the desktop.
Slashdot Mirror
Instead we're turning into a rabble of peasants and share-croppers slaving for, and kowtowing to, the modern day Lords of Corporatism. And we put on our chains so willingly!
The really sickening part is who owns all those corporations. It's the very same rabble of peasants who's being ground under foot.
The working class has far more total wealth than the upper class just because there's so many more working class people. That money is mostly held as corporate shares, either through CDs at the bank (the bank re-invests that money) or mutual funds.
It is the average person enslaving themselves here.
Jason
ProfQuotes
I still say the only way you had a 60% failure rate out of the box is if someone dropped the box. My HDs have almost always been out of those cases from tiny stores that buy bulk cases of parts intended for OEM computer assembly and sell them at retail. The only ones I've bought in retail packaging are when big box stores have mail in rebate specials that make them even cheaper.
Stop buying off eBay. I've bought about 10 Maxtor and Seagate IDE HDs for personal use in the past 5 years, and I've had 0 failures. I know that's pretty lucky (and I do keep carefuly backups), but come on, 60% failure right out of the box? Who dropped the box before you tried them?
:).
Of course now that I've said that, 60% of my drives will probably die in the next few days
Jason
ProfQuotes
No, my point all along is that it is essential for any good programmer to know how to program in assembly language, simply because it's necessary to really understand how the underlying hardware works so you can write effective code.
Your blind statement that your inefficient coding of what you're saying is an efficient algorithm is better than any possible efficient coding of a less efficient algorithm simply proves that point.
If you had a strong foundation in assembly, you wouldn't have made such a broad statement, because you'd understand the large demands quicksort has for system resources. Recursion is computationally very expensive to do, but in your high level ivory tower, that's all hidden from you. By choosing a platform that is incapable of the requried recursion, I just highlighted the point.
I'm also not sure what you're calling an analogy.
I never said a PIC controller was the ultimate computing platform. In fact, I picked a specific task the PIC was essentially useless for. It also is pretty useless at bit banging gigabit ethernet or USB. The baseline units don't even have an integer multiplication instruction.
I assume you'll be providing the fully compliant Javascript implementation for the PIC microcontroller?
You're the one who seemed to think Java was a good development choice. It should be your responsibility to make it work in the required environment.
Anyway, take a look at the Aino Java cross-compiler for the PIC family. It will compile Java to run on the PIC.
I'll take that bet...but since you choose the algorithms, I choose the architectures, and I choose a base-line PIC microcontroller. It has a 2-level deep hardware stack, Let's see your recursive javascript code run on that.
Jason
ProfQuotes
I use assembly every day to program embedded controllers. There are compilers for them, but I often use controllers with a few hundred instruction words of storage space and a few dozen bytes of ram. Compilers just aren't efficient enough.
Jason
ProfQuotes
This was non-military though it was confidential data (personal financial information about citizens).
At one time (and probably to this day) the US DOD specs used to require a certain number of passes of 0 and 1 bits followed by the writing of a specific bit pattern before a hard drive was considered to have been properly erased.
I find it hard to believe the US DoD is this lax on security. I used to work for the Canadian government, and we had to hammer a nail through the drive a certain number of times "according to the specs" to consider it properly erased.
Based on a binary floating-point number and switching system, it had all the attributes of today's computers, such as a control block, a memory, and a calculator. But it didn't have the ability to store the program in the memory together with the data because the memory was too small.
Modern computers don't necessarily have the program memory in the same space as the data memory. Machines using the Von Neumann architechture, such as a PC have a shared memory space. The newer Harvard architechture has separate program and data spaces.
There are many advantages to separating them. The main one is that you can concurrently fetch the data for the current instruction and the next instruction word from the two separate memory busses, effectively doubling your throughput. Also, you're going to want your memory to be a multiple of 8 bits wide, but there is no reason your instruction word should have this restriction.
Jason
ProfQuotes
He can't post pictures because's a BS post.
The is no mechanical throttle in an electric RC car to "tweak and superglue". It uses a PWM signal to an H-Bridge driver chip to the motor; all electronic connections. On a toy car, this is all done on one ASIC that gets RF in from the remote and outputs to the motor directly.
I've turned remote control cars in to various forms of robots; line following, sumo, etc.
It costs a fraction of the price of buying comparable parts from a normal supplier, and a lot of the mechanical work is done for you. Jason
ProfQuotes
If employees are willing to put in the hours, the company doesn't value the time anyway. If they cared about their employees, they wouldn't put such demands on them.
When I worked for a start-up, I was willing to put in extra hours as needed, but it was generally only needed to compensate for the gross mis-management of the company.
For example, we were developing a set-top video device, and there was only 1 test-model for the whole company. At one point, I needed to test some code on a wednesday morning, and my boss literally had me sit and watch for a chance to test it until friday evening. I wanted to do other work, but he explicitly said I was supposed to sit there and wait. On friday afternoon, he "authorized" me to come in on the weekend to do it, and acted like he was doing me a huge favour by letting me go in for no extra pay. Of course, I refused, but it also meant I was first on the chopping block when the company downsized a few months later.
When I was there, a typical work week was 70-80 hours (these people could have had a higher hourly wage at McDonalds), after the downsize, I kept in touch with some people there and it was closer to 100. 100 hours a week at a $50k/year (canadian) job; it's insane. It comes out to $10/hour for an exhausting an emotionally destructive job. Obviously these people have no life at all; the only people bringing dates/spouses to the christmas party was senior management.
Do what I do, and read the "old news" section instead of the front page.
That's a neat trick since this is still the most recent article and you've managed to post a comment on it long before it got to the old news section.
Jason
ProfQuotes
I hated that so much I cancelled my satellite TV service. With the money, I signed up with netflix.
So far it's great. Most of what I borrow are TV series DVDs, but I spend a lot less time watching TV and yet I enjoy it a lot more since I'm watching exactly what I want instead of just staring at the TV with "nothing" on.
As the AC said, I negotiate the best possible price before we even start talking extras. The dealers walk right into that because the purpose of the extras is to artifically raise the price so you're left feeling like you got a better deal than you really did. They don't want to mention the extras until you've agreed on a base price.
The "transportation fee" only applies if you're stupid enough to pay it. It's one of the easiest to get waived. The other is "administrative fee" which is basically their advertising budge. Getting those waived is as simple as saying you're not paying them and getting up to leave.
If Deutsch's books were written in the same style as that article, I don't think it's fair to credit Deutsch. The article is almost impossible to understand, and I'm very familiar with the experiment, it's written in such a condescending tone and with a high noise to signal ratio, it's difficult to distill out what it's trying to say.
Also, it pretends to be a home parallel universe test in the introduction, but in the meat of the article, it just dismisses the key point with no, this can't be done in your dining room, so it's hardly a home test at all.
Basically, this article takes an elegant concept in quantum physics, and presents it in an needlessly complex way without even truly explaining what is going on. It also pretends to be explaining how to do the key part of the experiment at home and then glosses over the fact that it's skipping that.
I'm surprised poking holes through paper with a pin even works, even for a pin hole camera, people use aluminum foil for the hole so it can be smaller and more precise.
Using a hole though isn't even the best way to do the experiment. It will cause difraction to occur in 2 dimensions, it is a lot easier to observe what's going on if it happens in 1 dimension. In high school, we blackened a microscope slide with smoke from a candle, then made 2 parallel scratches by taping 2 flat razor blades together and drawing them across the blackened glass. It gave nice perfectly paralell slits a known distance apart (the thickness of the blade).
Again this just demonstrated the first part, not the part Deutsch claims proves parallel universes exist.
Jason
ProfQuotes
This is a very old experiment, and a well-known phenomenon. It was even one of the answers on slashdot's poll for favourite physics experiment (and my personal favourite).
Even the idea that it is proof of parallel universes is not original. Michael Crichton made that claim in his book Timeline. It's an excellent book (despite the horrible movie loosely based on it), but it is fiction.
Jason
ProfQuotes
Isn't that Sony's motto these days?
I have had many a customer tell me that NO ONE has ever opened their computer (including them) - I open it up and there are screws missing, the magnetic sheild has fingerprints on it, etc etc.
Interesting, because I bought my Wallstreet brand new, it never had been opened by anyone (including me), until I opened it a few months ago to replace the sound board. There were no missing screws, but there were fingerprints on the shield over the CPU.
I have enough pc and notebook repair experience that I don't allow "service providers" to touch my computer; they're primarily concerned with speed, not quality. I've seen machines come back from "authorized" service centers with stripped or missing screws and even broken plastic parts; the hinge cluth and upper case plastic on the Wallstreet, are a pain to take off for example, it's quite common for things like that to get broken at the service center.
Of course the service tech doesn't say anything, leaves the broken part in, and even charges full price. I hate taking my computers in for repair under warranty, I would prefer to just get the replacement part and install it myself, but of course they won't allow that.
If I've added an upgrade to my notebook, why should I lose the warranty when an unrelated part fails due to a manufacturing or design defect within the warranty period?
Jason
ProfQuotes
The only difficulty I had was keeping track of all the screws.
I took apart my wallstreet to fix the faulty power connector on the sound card (it's a common defect, apple's design flaw). My solution to keeping track of the screws was to use my digital camera and take lots of pictures. Usually I'd place the screws beside the laptop in the same relative positions they belong in.
Also, I found This site very useful as a disassembly guide. The site posted is hardly a great new idea.
Jason
ProfQuotes
If you did anything with those drives (load a program, just leave it powered of, etc), the heads would fall out of alignment. It was fairly simple regular maintainence to realign the drive. The motor even had that sticker for the stobe freeze effect under 60Hz fluorescent lighting.
Jason
ProfQuotes
Any linux expert would have the network card working inless than 5 minutes in any of the development oriented distros.
Exactly my point. The OP said this is the year of linux, I said it isn't. It seems that you agree with me.
Your reply is kind of funny, you are blindly defending linux and attacking me for saying it's not the greatest thing ever invented, yet at the same time your attacks prove me point. If linux is not for me (with a CS degree), it sure isn't for most users.
I switched from Linux to Windows about 3 years ago. It was just a pain in the ass to try and do anything in linux. You want to install a network card? Better set aside a weekend to recompile the kernel and then spend countless frustrating hours trying to configure everything.
Plug the same card into a windows machine and you're browsing the web in under 5 minutes.
Just this week I'm trying to give linux another shot. I have an old powerbook I want to set up as a wireless web browsing station, and WiFi cards that it supports are all 5 year old tech which is still a lot more expensive than a new card, so using YellowDog linux seemed like an ideal solution. The main install went well enough, expect it's only got a 2 gig drive and the distro puts a lot of useless junk on (like PPoE) so there's no room for X.
Then I have to spend hours browsing poorly written howtos and websites to find a driver that *might* support my network card. I downloaded the prism2 sourcefiles (since it's running on a PPC, I couldn't find a binary version), and then when I try to 'make configure' it wants the kernel source code...okay, I download the kernel source code, and prism still says it's missing or incomplete. I even tried recompiling that source and I get the same problem.
It's just not worth it. I'll spend the $1600 on a new powerbook (that will run OSX and includes an apple WiFi card). Anything is better that continuing to struggle with linux.
This is not the year of linux yet. Maybe I'll try it again on a desktop in a couple of years.
It works well enough on the server; I do have a linux server that started as slackware 3.3, and maintaining server features is reasonably easy, but linux is still useless on the desktop.