While it's probably legally unenforcable, I'd love to see this. All the lingering bugs in my Windows video drivers, gone.. I might actually go a week without a crash, imagine that.
What about a noisy consumer group, instead of or in addition to a law? A Consumer Reports for computers, that comes back and does yearly re-reviews of old products. Tie them to a legal group that would support class-action lawsuits on behalf of abandoned customers, and we might have something.
IANAL, but... Auto manufacturers who make exploding fuel tanks FIX them, or they get sued, and they lose! Guns that blow up in people's faces, when loaded with only the ammunition they're rated for, ARE behind many lawsuits. There's a difference between user stupidity and faulty design. In the real world, companies are held liable for grossly negligent design. It's just computers, which are toys and can cause no real damage, that are exempt from these laws.
Now consider this: If the US Government thinks that computers are now an important enough part of our economy to justify all these infrastructure task forces and so on, won't it eventually occur to someone that computers aren't toys any more, that a significant chunk of our economy depends on the proper operation of these machines, and that careful software design is essential to all of it?
If all those taxpayer-funded task forces wanted to do something useful, they'd come out and publicly proclaim one simple fact: "Commercial software makers should be just as liable for negligence as commercial automobile/gun/etc makers." Suddenly, with big bucks on the line, I think we'd see holes like this get fixed or not opened in the first place.
Microsoft's arrogance towards security comes from their EULAs. Those licenses, in turn, only make sense if computers are toys. Out here in the real world, where a [car/computer] that crashes can cause damage to your [forehead/business], the designer of the faulty [spring/program] can be held liable if the failure was predictable/obvious/forseeable.
I am NOT a lawyer, this is just based on my understanding of the status quo. I'd love to hear from real lawyers who can back me up or slap me around on these points.
I was doing this under DesQview/386 with a VGA and a CGA card sitting on an ISA bus. Since they don't share the same video memory addresses, it was easy. I used the VGA for anything pretty, and ran a text console on the CGA, with a little window open on the side showing my memory stats, system load, and modem indicators.
I don't understand it.. Mir is falling apart, very risky, and expensive to maintain. And how much profit are they really gonna make on each visit? Whereas Iridium takes what, a couple guys with telescopes and slide-rules to make sure the birds stay in the right orbits? You sure don't need much of an insurance policy with the thing, and I'm sure Iridium's user base could generate more money than a couple high-priced stays in the Mirtel 6. (We'll leave the oxygen on for ya.)
Sure, most telco transport equipment lets you "drop-and-continue" a signal. The logistics would be difficult, satellites are really better for broadcasting, where latency isn't an issue.
And if you don't know what WDM means, do some research before making another irrelevant post.
Sure, you could use fiber for the interconnects within a computer. But why? The signal starts as electrical, and it needs to be processed by more electronics which are only a few inches away. The expense, heat, and latency of converting it to photons is absurd when compared to simply sending it over copper.
Now when you start getting into external perhipherals, fiber becomes viable. The expense of making a cable and connectors that can withstand life outside your case, transmit lots of signal very quickly, and perhaps get a signal several meters down the room, fiber suddenly becomes a really good idea.
However, even when we're talking about Fibre Channel SCSI, we're still referring to multimode, which is way different from the fiber that telcos use. I'm not going to go into the details, but it has to do with the way light pulses stretch out as they pass through the fiber.
These researchers are pushing limits that have nothing to do with your garden-variety optical semaphore. The speed at which your SCSI controller might modulate light will result in pulses several meters long (do the math -- speed of light, frequency of signal) so it doesn't much matter what color they are, and you can send several colors down a fiber with no problem. Now try sending pulses so fast that each pulse is only a couple waves long. Suddenly it becomes more difficult to cram closely-spaced colors onto the same piece of glass.
Are you beginning to get a sense of why multiterabit optical links aren't practical inside your computer?
So then you'd need to tell the equipment which lines to protect and which ones to ignore if they fail? That's quite impossible to do. Consider a card that serves 672 customers. What do you do when it fails? Chances are, one of those customers is going to be paying for POTS, not SPLAT service.. (Simple Plain Low Availability Telephone) So you have to protect that card!
Alternately, you'd need to set up a 100%-separate system, built out of cheap unrpotected parts. Then when someone requests a change to cheap service, you incur labor when you move their jumper to the appropriate system. Not to mention the expense of installing the SPLAT system initially.
By the way, as companies reorganize and merge, they dump their good people. This is a time-proven fact. Untrained people end up responsible for maintaining these systems, and things just don't go right. On condition of anonymity, I can quote an employee of the local telco:
"The company can lie to the FCC, they can lie to the state regulatory board, they can lie to the customers. But they can't keep anything from their employees. We know what kind of service the company could be providing. Many of us have been around long enough to know how good it can be. Trust me, if the customers knew the kind of piss-poor service they're really getting, they'd be rioting outside company headquarters."
Hey, I've hot-plugged ISA cards before.. Just make sure no software is expecting to use it, the IRQ is clear, engage the ground contacts first, and cross your fingers.
It really freaks PC people out when you yank cards out of a working system, doesn't it? During the course of testing a system, we beat on the chassis and cards with our fists, wiggle and pull on the connectors, bodycheck the entire bay of equipment, pull each card, reset the processor, interrupt any signals that are supposed to be protected, and if at any time traffic fails for more than 50ms, we have to go back and figure out what went wrong. Computer users just don't understand it.
It's difficult, to say the least, to get people to understand that when we apply power to something, it never goes down, for any reason, until it's obsolete and ready to be removed.
Do you have an example, or some good way of explaining this sort of stuff to people outside the industry? I seem to get blank stares when I describe telco availability standards to PC people.
I've been waiting for this... TTC has been using Solaris in their Centest offerings for a while, and their TestPad 2000 series products actually run DOS and Windows(!). These are only test instruments, so their accuracy and ease of use are more important than uptime; customer traffic isn't affected.
Nortel runs HP-UX in some of their transport equipment, but again it's a non-service-affecting application. Failure of the overhead processor means that performance monitoring and protection switching are lost, but it doesn't immediately affect traffic. I don't know what the DMS-series switches run at the core, but the user interface looks the same as on their TransportNodes.
Tellabs runs their Titan series cross-connect systems on PowerPC processors. As in the Nortel equipment, the traffic itself is carried on dumb electronics; loss of the processor only affects fault recovery, system provisioning, and performance monitoring.
So far, nobody's using Linux for mission-critical stuff, processing customer calls in real-time. This is probably about to change! Slashdot readers know that Linux is more stable than the average desktop OS. But most people don't realize the extreme requirements of the telecom industry.
For instance: When a tornado ripped the roof off a central office and half the switch was soaked, the parts which weren't physically destroyed by water kept running.
This is an industry where there's (hopefully) no such thing as downtime. I've been in offices where data circuits have been functioning continuously since before I was born. A few bit errors here and there due to the occasional lightning strike, but no real interruptions. From the switches that actually handle your calls, to the transport systems that move data from one office to another, everything has backups. Commercial power fails? No problem, the office runs on batteries anyway. They go from charging to discharging, and you've got 12 hours to get the diesel generator running in case it doesn't start itself. After that, you've got a week's worth of fuel in an underground tank. Let's say some knucklehead throws a wrench into a power board. Instant pinkslip, but the customers never know, because everything has two power feeds. Down to the individual card level, every circuit in a piece of telcom equipment has a backup that takes over in the event of a failure.
In the PC world, RAID comes close to this level of reliability in terms of a drive failure, but how many of them can give you access to your data even if a controller or bus fails?
Is your desktop box ready for this?
HA Linux IS a significant development. I haven't had a chance to check out the specs yet, (Slashdotted -- how's that for availability?) but from the quick blurb here, I can say that this will seriously change some things in the carrier market. Your ESS or DMS or EWSD might not run Linux any time soon, but some enormous routers and call-processing systems might.
A Commodore 64 emulator would be even better. Imagine it, buy a new game console and have instant access to all your favorite games of yore. THAT would boost sales.
I'm sure that Slashdotters will make up a large portion of the votes received, when the process happens. I'm picturing something not unlike the Interview process, where moderators pick the posters who seem to know what they're talking about, and then the top ten or so are selected.
No, but it makes sense to me that all my AOLish friends might become puppet members. "Hey, guys, register yourselves here. I'll tell you how to vote, and I'll make it worth your while."
The ones that actually fit in business card holders are 50MB. The slightly larger ones, which you can still wedge into your wallet, are 140. A plain round 3" is 200 or so, I think. 250 perhaps? Then vanilla 5"ers are 650 and you can get overburn-ready discs that go to 700 and if your drive is capable of it, some of these can go to about 708. Check www.ahead.de and look at Nero, everybody's favorite Win9x-based CD authoring software, which includes overburn support.
It's like Renault Encore owners carry a spare transmission in the trunk. You're responsible for making whatever adaptations are necessary to compensate for the shortcomings of your hardware.
Shouldn't those slot drives ship with an adapter ring or two?
When 3" CD-Audio singles came out, many of them came with a small adapter which was basically a hollow 5" cd with some sort of micro clip mechanism to hold the mini in the middle.
I've never used one, and I think they might have required the mini to have a shaped edge, so that the clip wouldn't be thicker than the CD. So they might not be compatible with the mini CD-Rs and rectangular minis we're seeing now.
If anyone has one of these things, I'd like to hear about whether it works.
When they make a real CD, they wash the media layer away from the outer millimeter before applying the coating. This allows the polycarbonate shell to be continuous around both sides of the CD, and air never touches the media layer.
If you cut your own and get the balance perfect, it'll work for a while. But who knows what oxygen might do to the CD-RW dye? The edges might become unusable after a while, and I don't know how fast it would spread.
I don't know the chemical makeup of the dye, but I'm sure someone could tell you why they go through the bother in the first place of keeping it away from air at the edge. cd-recordable.com used to have a wonderful section showing how CD-Rs are made, but they seem to have changed affiliations and that section disappeared.
Yup, he was convinced that the distribution medium was what made his scheme great and gave it the means to succeed where so many other commerce systems had failed.
I just kept him around long enough to get some pictures of the CD, since I'd never seen one before. He seemed amused by my quickcam too.
They're about the same price here at the swap'n'swindle.
My plans for it include a disc with the Windows distributed.net client, an autorun.inf and a bootable section. The autorun and the autoexec will both install the client. Simply inserting the disc into a Windows machine will contribute to my keyrate.;)
Shhh....
My rescue CD is way too big to fit on one of these things. It nearly fills a 5" CD-R now.
"Had it been done right" stated the original post.
I take this to mean "If multiple processors were simpler than present, if motherboards and architectures and OSs were easily scalable".
All the stuff you mention: Cache glue, software spaghetti, cooling and power.. is based on the assumption that current techniques would be adapted to suit multiprocessor systems. Well, no, current techniques are best-suited to single processors. That's why we're here.
But what if there'd been some major breakthrough, some innovation that allowed multiprocessing to become simple? Perhaps a way to put intelligence in compilers so that programs could run, segmented across many processors and in a divided memory space. Perhaps a way for lots of processors or mini-machines to communicate efficiently, possibly adding redundancy for error-checking. Maybe something else, or a combination of the above.
Then what we'd see is less emphasis on making faster individual chips, and more on making the chips cooperate more effectively. We'd still be making faster chips, because that is one way to get more performance out of a system. But we wouldn't be screaming along near the gigahertz mark, because it simply wouldn't be necessary.
Multiprocessing wouldn't replace development of faster chips, it would supplement it. We'd be a lot better off with a couple dozen 200MHz chips than a single 800MHz, and we'd be a lot more tolerant if one CPU fan failed.
Are they including distributed.net as a default package? Hehe.. If it takes effort to turn the thing off, someone's keyrate would skyrocket. (or even be nice about it, and set it up with the user's own email)
Slashdot Mirror
c:\music\misc\NotMetallica-TheSmurfSong.mp33
c:\music\misc\NotMetallica-InspectorGadget.mp3
c:\music\misc\Morning_mix_Metallica-Sucks-Me.mp
Anyone?
While it's probably legally unenforcable, I'd love to see this. All the lingering bugs in my Windows video drivers, gone.. I might actually go a week without a crash, imagine that.
What about a noisy consumer group, instead of or in addition to a law? A Consumer Reports for computers, that comes back and does yearly re-reviews of old products. Tie them to a legal group that would support class-action lawsuits on behalf of abandoned customers, and we might have something.
IANAL, but... Auto manufacturers who make exploding fuel tanks FIX them, or they get sued, and they lose! Guns that blow up in people's faces, when loaded with only the ammunition they're rated for, ARE behind many lawsuits. There's a difference between user stupidity and faulty design. In the real world, companies are held liable for grossly negligent design. It's just computers, which are toys and can cause no real damage, that are exempt from these laws.
Now consider this: If the US Government thinks that computers are now an important enough part of our economy to justify all these infrastructure task forces and so on, won't it eventually occur to someone that computers aren't toys any more, that a significant chunk of our economy depends on the proper operation of these machines, and that careful software design is essential to all of it?
If all those taxpayer-funded task forces wanted to do something useful, they'd come out and publicly proclaim one simple fact: "Commercial software makers should be just as liable for negligence as commercial automobile/gun/etc makers." Suddenly, with big bucks on the line, I think we'd see holes like this get fixed or not opened in the first place.
Microsoft's arrogance towards security comes from their EULAs. Those licenses, in turn, only make sense if computers are toys. Out here in the real world, where a [car/computer] that crashes can cause damage to your [forehead/business], the designer of the faulty [spring/program] can be held liable if the failure was predictable/obvious/forseeable.
I am NOT a lawyer, this is just based on my understanding of the status quo. I'd love to hear from real lawyers who can back me up or slap me around on these points.
I was doing this under DesQview/386 with a VGA and a CGA card sitting on an ISA bus. Since they don't share the same video memory addresses, it was easy. I used the VGA for anything pretty, and ran a text console on the CGA, with a little window open on the side showing my memory stats, system load, and modem indicators.
Ha!
I don't understand it.. Mir is falling apart, very risky, and expensive to maintain. And how much profit are they really gonna make on each visit? Whereas Iridium takes what, a couple guys with telescopes and slide-rules to make sure the birds stay in the right orbits? You sure don't need much of an insurance policy with the thing, and I'm sure Iridium's user base could generate more money than a couple high-priced stays in the Mirtel 6. (We'll leave the oxygen on for ya.)
Sure, most telco transport equipment lets you "drop-and-continue" a signal. The logistics would be difficult, satellites are really better for broadcasting, where latency isn't an issue.
And if you don't know what WDM means, do some research before making another irrelevant post.
Sure, you could use fiber for the interconnects within a computer. But why? The signal starts as electrical, and it needs to be processed by more electronics which are only a few inches away. The expense, heat, and latency of converting it to photons is absurd when compared to simply sending it over copper.
Now when you start getting into external perhipherals, fiber becomes viable. The expense of making a cable and connectors that can withstand life outside your case, transmit lots of signal very quickly, and perhaps get a signal several meters down the room, fiber suddenly becomes a really good idea.
However, even when we're talking about Fibre Channel SCSI, we're still referring to multimode, which is way different from the fiber that telcos use. I'm not going to go into the details, but it has to do with the way light pulses stretch out as they pass through the fiber.
These researchers are pushing limits that have nothing to do with your garden-variety optical semaphore. The speed at which your SCSI controller might modulate light will result in pulses several meters long (do the math -- speed of light, frequency of signal) so it doesn't much matter what color they are, and you can send several colors down a fiber with no problem. Now try sending pulses so fast that each pulse is only a couple waves long. Suddenly it becomes more difficult to cram closely-spaced colors onto the same piece of glass.
Are you beginning to get a sense of why multiterabit optical links aren't practical inside your computer?
So then you'd need to tell the equipment which lines to protect and which ones to ignore if they fail? That's quite impossible to do. Consider a card that serves 672 customers. What do you do when it fails? Chances are, one of those customers is going to be paying for POTS, not SPLAT service.. (Simple Plain Low Availability Telephone) So you have to protect that card!
Alternately, you'd need to set up a 100%-separate system, built out of cheap unrpotected parts. Then when someone requests a change to cheap service, you incur labor when you move their jumper to the appropriate system. Not to mention the expense of installing the SPLAT system initially.
By the way, as companies reorganize and merge, they dump their good people. This is a time-proven fact. Untrained people end up responsible for maintaining these systems, and things just don't go right. On condition of anonymity, I can quote an employee of the local telco:
"The company can lie to the FCC, they can lie to the state regulatory board, they can lie to the customers. But they can't keep anything from their employees. We know what kind of service the company could be providing. Many of us have been around long enough to know how good it can be. Trust me, if the customers knew the kind of piss-poor service they're really getting, they'd be rioting outside company headquarters."
Hey, I've hot-plugged ISA cards before.. Just make sure no software is expecting to use it, the IRQ is clear, engage the ground contacts first, and cross your fingers.
It really freaks PC people out when you yank cards out of a working system, doesn't it? During the course of testing a system, we beat on the chassis and cards with our fists, wiggle and pull on the connectors, bodycheck the entire bay of equipment, pull each card, reset the processor, interrupt any signals that are supposed to be protected, and if at any time traffic fails for more than 50ms, we have to go back and figure out what went wrong. Computer users just don't understand it.
It's difficult, to say the least, to get people to understand that when we apply power to something, it never goes down, for any reason, until it's obsolete and ready to be removed.
Do you have an example, or some good way of explaining this sort of stuff to people outside the industry? I seem to get blank stares when I describe telco availability standards to PC people.
I've been waiting for this... TTC has been using Solaris in their Centest offerings for a while, and their TestPad 2000 series products actually run DOS and Windows(!). These are only test instruments, so their accuracy and ease of use are more important than uptime; customer traffic isn't affected.
Nortel runs HP-UX in some of their transport equipment, but again it's a non-service-affecting application. Failure of the overhead processor means that performance monitoring and protection switching are lost, but it doesn't immediately affect traffic. I don't know what the DMS-series switches run at the core, but the user interface looks the same as on their TransportNodes.
Tellabs runs their Titan series cross-connect systems on PowerPC processors. As in the Nortel equipment, the traffic itself is carried on dumb electronics; loss of the processor only affects fault recovery, system provisioning, and performance monitoring.
So far, nobody's using Linux for mission-critical stuff, processing customer calls in real-time. This is probably about to change! Slashdot readers know that Linux is more stable than the average desktop OS. But most people don't realize the extreme requirements of the telecom industry.
For instance: When a tornado ripped the roof off a central office and half the switch was soaked, the parts which weren't physically destroyed by water kept running.
This is an industry where there's (hopefully) no such thing as downtime. I've been in offices where data circuits have been functioning continuously since before I was born. A few bit errors here and there due to the occasional lightning strike, but no real interruptions. From the switches that actually handle your calls, to the transport systems that move data from one office to another, everything has backups. Commercial power fails? No problem, the office runs on batteries anyway. They go from charging to discharging, and you've got 12 hours to get the diesel generator running in case it doesn't start itself. After that, you've got a week's worth of fuel in an underground tank. Let's say some knucklehead throws a wrench into a power board. Instant pinkslip, but the customers never know, because everything has two power feeds. Down to the individual card level, every circuit in a piece of telcom equipment has a backup that takes over in the event of a failure.
In the PC world, RAID comes close to this level of reliability in terms of a drive failure, but how many of them can give you access to your data even if a controller or bus fails?
Is your desktop box ready for this?
HA Linux IS a significant development. I haven't had a chance to check out the specs yet, (Slashdotted -- how's that for availability?) but from the quick blurb here, I can say that this will seriously change some things in the carrier market. Your ESS or DMS or EWSD might not run Linux any time soon, but some enormous routers and call-processing systems might.
Linux for one of these machines would be cool..
A Commodore 64 emulator would be even better. Imagine it, buy a new game console and have instant access to all your favorite games of yore. THAT would boost sales.
I'm sure that Slashdotters will make up a large portion of the votes received, when the process happens. I'm picturing something not unlike the Interview process, where moderators pick the posters who seem to know what they're talking about, and then the top ten or so are selected.
No, but it makes sense to me that all my AOLish friends might become puppet members. "Hey, guys, register yourselves here. I'll tell you how to vote, and I'll make it worth your while."
*boink* That's the sound of my brain going "wow!".
Cool, thanks!
The guys at cd-recordable.com have been making colored, black, and aryan CD-Rs for quite some time. They go for about $2 a disk in bundles of ten.
So when are we going to see colored-plastic business card cut CD-RW discs?
The ones that actually fit in business card holders are 50MB. The slightly larger ones, which you can still wedge into your wallet, are 140. A plain round 3" is 200 or so, I think. 250 perhaps? Then vanilla 5"ers are 650 and you can get overburn-ready discs that go to 700 and if your drive is capable of it, some of these can go to about 708. Check www.ahead.de and look at Nero, everybody's favorite Win9x-based CD authoring software, which includes overburn support.
It's like Renault Encore owners carry a spare transmission in the trunk. You're responsible for making whatever adaptations are necessary to compensate for the shortcomings of your hardware.
Shouldn't those slot drives ship with an adapter ring or two?
When 3" CD-Audio singles came out, many of them came with a small adapter which was basically a hollow 5" cd with some sort of micro clip mechanism to hold the mini in the middle.
I've never used one, and I think they might have required the mini to have a shaped edge, so that the clip wouldn't be thicker than the CD. So they might not be compatible with the mini CD-Rs and rectangular minis we're seeing now.
If anyone has one of these things, I'd like to hear about whether it works.
When they make a real CD, they wash the media layer away from the outer millimeter before applying the coating. This allows the polycarbonate shell to be continuous around both sides of the CD, and air never touches the media layer.
If you cut your own and get the balance perfect, it'll work for a while. But who knows what oxygen might do to the CD-RW dye? The edges might become unusable after a while, and I don't know how fast it would spread.
I don't know the chemical makeup of the dye, but I'm sure someone could tell you why they go through the bother in the first place of keeping it away from air at the edge. cd-recordable.com used to have a wonderful section showing how CD-Rs are made, but they seem to have changed affiliations and that section disappeared.
Yup, he was convinced that the distribution medium was what made his scheme great and gave it the means to succeed where so many other commerce systems had failed.
I just kept him around long enough to get some pictures of the CD, since I'd never seen one before. He seemed amused by my quickcam too.
They're about the same price here at the swap'n'swindle.
;)
My plans for it include a disc with the Windows distributed.net client, an autorun.inf and a bootable section. The autorun and the autoexec will both install the client. Simply inserting the disc into a Windows machine will contribute to my keyrate.
Shhh....
My rescue CD is way too big to fit on one of these things. It nearly fills a 5" CD-R now.
Yes, but how dense is this stuff?
*cough*microsoft*cough*
"Had it been done right" stated the original post.
I take this to mean "If multiple processors were simpler than present, if motherboards and architectures and OSs were easily scalable".
All the stuff you mention: Cache glue, software spaghetti, cooling and power.. is based on the assumption that current techniques would be adapted to suit multiprocessor systems. Well, no, current techniques are best-suited to single processors. That's why we're here.
But what if there'd been some major breakthrough, some innovation that allowed multiprocessing to become simple? Perhaps a way to put intelligence in compilers so that programs could run, segmented across many processors and in a divided memory space. Perhaps a way for lots of processors or mini-machines to communicate efficiently, possibly adding redundancy for error-checking. Maybe something else, or a combination of the above.
Then what we'd see is less emphasis on making faster individual chips, and more on making the chips cooperate more effectively. We'd still be making faster chips, because that is one way to get more performance out of a system. But we wouldn't be screaming along near the gigahertz mark, because it simply wouldn't be necessary.
Multiprocessing wouldn't replace development of faster chips, it would supplement it. We'd be a lot better off with a couple dozen 200MHz chips than a single 800MHz, and we'd be a lot more tolerant if one CPU fan failed.
Are they including distributed.net as a default package? Hehe.. If it takes effort to turn the thing off, someone's keyrate would skyrocket. (or even be nice about it, and set it up with the user's own email)