I wasn't aware they still were making that stuff. It would also be neat to see some original work in larger technologies, specifically tailored to the super-reliability market. I guess it would be too expensive. Your remark about 'lots of atoms to knock out of several microns' is exactly on the mark, whether the knocking is being done by radiation, phonons, or electrons of normal operation. But in today's technologies, just about every structure is submicron.
I agree with you 100%, and have long felt that the NRA is the right example. More appropriate than many might thing, even. After all, crypto is a big geek issue, so we are very much into, "The right to bear arms." But getting too close to the NRA might make for strange bedfellows. (except for ESR, of course)
Is it time for the Geek community to target...
on
MPAA vs. Television
·
· Score: 5, Interesting
People like Hollings for non-re-election? Perhaps we need "The Geek Lobby Page" where information about key publicly elected officials is kept.
When is Hollings up for re-election?
Who is running against him?
Are the opponents views any better?
We all grumble, complain, and flame. We also say we're too small. But have we tried yet to use tried-and-true mainstream political techniques?
But I believe that long-term space operation is just a place where current electronics haven't ever gone before. Current electronics are obsolete and in the landfill well before hitting the wearout limits. In truth, I believe relatively little has been done about investigating true wearout compared to the quality control issues you mention.
But IMHO current electronics simply won't last the way older, larger geometries did. OTOH, if we were to take today's quality capabilities and knowledge, and build some appropriately larger geometries, we truly could build long-duration space electronics.
I seriously question the long-term of any semiconductor electronics built today. No, there are no moving parts - except the electrons and any atoms they may knock about as they scurry on their way from source to drain and through the wires.
Shipping reliable semiconductors has always been a lifetime issue. There is a "bathtub curve" of failures, with a higher number of early fallout, then a very reliable main lifetime, then failures rise again at wearout. Wearout happens through mechanisms like electromigration, where the electrons physically knock the metalization atoms out of place. In addition, all of the hot process steps like diffusion continue to happen, just at much slower rates. High reliability semiconductors are "burned in", run at higher temperatures and voltages than normal, to force them past that early fallout and throw those parts away.
So what does this mean to space electronics? First, radiation just doesn't help. You can design rad-hard, but the crystal lattice is still taking damage, and it's cumulative. The low temperature helps to slow down wearout mechanisms.
But the big problem is modern technology. The smaller geometries will simply wear out faster. Finer wires are more subject to electromigration, though using copper is an improvement because the atoms are heavier than aluminum. But gates are thinner, as are diffusions and spacings, non of which helps long life. When designing a burn-in regimen, it's getting tougher to get past early failures without approaching wearout. While frequency can be reduced to increase lifetime, scaling voltage down is getting tougher, because we're running darned close to minimums, already.
One of my pet thoughts is the idea of electronics for a multi-generation starship. Other than slowing it down, stopping as much as possible, reducing voltages, etc, it's a tough problem. Maybe the best way is to scrape the bargain bins for old technology.
Nor am I terribly serious about photography, but my family and I do like our vacation slides. We used to use the Nikon FA for slides, and the FT2 for snapshots, but the latter has been waiting for a round tuit for too many years to get the light meter fixed. So the FA has been swapping between print and slide film, unconvenient at times.
In the medium run, I see moving the FA back to slides, and getting a digital for snapshots. That leaves me with a serious controllable camera when it really matters, and pre-selection on snapshots when that matters.
In the longer term, I don't know what we'll do. I have a small collection of Nikon mount lenses, though I guess they're all obsolete by any contemporary terms. Certainly predates autofocus. I wish someone would come up with a digital SLR back too, and upgradable CCDs would sure be a good idea, while they're at it.
But perhaps the SLR/prism viewer itself is obsolete, given that the CCD can deliver a preview image just fine. But the whole eyepiece viewing and control is certainly more precise than a mini-preview screen. Perhaps if they could deliver the image through an eyepiece and keep the better muscular control.
at least when I went there. aka See double, you are you.
Are the Spitwad and the Fountain still in the quad? Is Presti's Donuts still in business, and are the Best Cinnamon Rolls in the known universe still available there at 1:15 AM?
In that case, read "Sundiver" by David Brin. Good book, first of the first Uplift Trilogy. (Though IMHO the second book, "Startide Rising" was better, and the third, "The Uplift War" was somewhat weak.
This whole thing seems familiar, not just with GIF but with Rambus and the SDRAM/DDR standards.
At the time, one of the writings mentioned a thing called "Equitable Estoppel." My interpretation was that if you had a patent that was becoming an industry standard, you had to begin notification "promptly," and to allow it to become a standard and *then* begin notifying/litigating was legally naughty.
Rambus is still around, though a shadow of their former arrogance. (I understand that the people are still just as arrogant as ever, they just don't get the press.) In some ways, notably submarining and patent-stretching Rambus was worse. But at least once they had stretched their original art to look like it covered SDRAM and got it issued, they were prompt in filing suit.
>I used to really believe in total free-market capitalism
I suspect you and a lot of others are missing the point here, though you've eventually come to the right conclusion.
People are not capitalists, the collective is. Capitalism is not based on people believing in it, it's based on people being greedy. Government oversight is *necessary at the theoretical level* for capitalism because the participants need not believe in it, and the most successful participants generally don't. Pretty much every successful business' goal is to corner the market and sit back collecting money. No business has as its long-term goal to engage in continuous cutthroat competition with its peers. No business has as its long-term goal to reduce prices and improve value for the customers. Those are not long-term goals, those are necessary means to the real long-term goal - making money.
Nothing wrong with this. Capitalism simply recognizes that greed is an effective motivator. But it requires recognition that an outside regulating force is necessary in order to prevent the end Marx predicted.
Unfortunately in the past decade or so, people have come to worship the free market without realizing its inherent instability, and have worked to remove the external corrective forces.
Others have mentioned Microsoft's giant cash pile, and said that they can keep giving away Windows practically forever in the third world.
But it doesn't matter, this HURTS.
First off, there's the simple fact that Linux made it onto the table, and it took Bill Gates himself to go 'bring back the account.' That's quite an endorsement Microsoft has given Linux, right there.
Second, Linux made it onto the table, here and in Norway. Simply getting onto the table is new. Maybe for the next few rounds Linux will get onto the table as a bargaining ploy. But one of these times it's going to be a WIN.
Third, Microsoft is sitting on a big pile of cash, just about everyone knows that. But now a big part of their invincible image has become that pile of cash. It's no longer completely fluid, it has become an essential holding. Besides, the death of a thousand paper cuts is perhaps more likely than a big wound. So add up X-Box subsidies, third-world givaways, European givaways, (Norway, anyone?) and no doubt some 'License V6 adjustments' to keep customers in the fold, and it's going to be tough to keep that wad of cash.
That cash is also only part of the image. The Gates trip to Peru chips away, License V6 chips away, the DOJ trial took a big chunk out, etc.
I can't wait for hordes of Elvis fans to protest against the title of, "Return of the King," or will the be in the theatres apoplectic when they find that the king who returned *wasn't* Elvis.
Or since there have been any number of treatises examining Lord of the Rings for Christian elements and Nuclear warnings, etc... How about evidence that JRRT could see the future and named the book accordingly?
Otherwise, Ms. Portman, naked and petrified, played bit parts in numerous scenes. Most notable was her 'falling rock' part in the stair scene during the Balrog attack. Some of her other scenes were largely covered by vegetation, so I guess they would only qualify as petrified.
Actually there are secure extensions to both telnet and smtp. The security problem can be separated from the protocol itself, and there are several such implementations. By the same token, with ssh and ssl/tls the same exists. Agreed it's a little tougher to see, once the bytes get encrypted, but the scheme is carefully documented and can be traced *by the authorized parties*.
Actually, FWIU there *is* something wrong with.doc. Again, this is all gleanings, but to sally forth into the darkness:
"Good" protocols are things like telnet, smtp, etc. They are simple, straightforward, and discoverable. It seems that.doc has none of those three attributes. If it did, someone would have created really good.doc import/export filters by now, not just almost-decent ones. There are some stories about that.doc is not even documented inside Microsoft, rather that there is a 'reference implementation' of source code.
A "good" file format, from a technical perspective, would have offered much better compatibility between revisions. The.doc format appears to be engineered largely to force users to upgrade.
I'd say that there are some intrinsic problems with.doc.
The current administration has not been very friendly to such highly regulated non-business-friendly things as a public IP network. The concept of 'public good' seems to have been tossed out the window. OTOH, the current round of business scandals certainly sets the tone for such a thing. Several of the current scandals threaten to darken serious portions of the Internet, at least temporarily. My own home connection (Adelphia) is threatened, for that matter.
On the more historic side, apparently nobody remembers The Source or any of the early highly-proprietary online services. Even CompuServe, AOL, and Prodigy only survived by becoming Internet portals, though all but AOL are all but gone. People simply didn't want to be locked in. They wanted the "IP Utility" that the Internet originally offered. Ever since the Internet was privatized, there's been a tug-of-war to turn it back into a proprietary cash-cow, despite the teachings of even recent history.
But then again, we went to the Moon, and threw all of that away.
Yeah, Hiro did have Reason, but the sword was in the Metaverse, not reality, and he did hack his way through at least some of the problems. Uncle Enzo is allowed to use more physical things, because that's where his power was rooted - he wasn't a hacker.
But this one was a bit more fisticufish than some other Stephenson. I was thinking more of Cryptonomicon or Diamond Age. A little of the hero doing fisticuffs there too, but still reasonable.
I seem to remember hearing that some of the German rocketry work was done by Jewish slave labor under inhuman conditions. Don't know if it was as bad as the medical experimentation, though.
IMHO, the secondary damage of the War On Drugs far outweight the damage of the drugs, themselves. By trying to interdict drugs, we turn them into a high-dollar business. Also IMHO, it becomes self-defeating, because the more successful you are at interdiction, the more the cost rises, the better-financed the drug supply chain become.
I heard on NPR that in the 1968 election, Nixon promised to reduce crime. After getting elected, he had to produce. Some of his advisors told him that all of the more traditional law enforcement techniques had been proven not to work. (by experience) His best shot was drug treatment, to reduce demand, and therefore crimes of financing. He went along with it, it worked, and crime actually did go down measurably. In the 1972 election crime was no longer a big issue, so he dismantled the apparatus, and the approach has never been taken seriously again.
Kind of like the way Clinton/Greenspan actually did achieve a "soft landing" of the economy right before the dot-com boom. But now the concept appears to be forgotten, and I have no doubt that whenever recovery comes, it will be back into the usual boom/bust cycles.
Back to topic, filesharing is an interesting comparison to drugs because it is a widespread crime. Perhaps it should be better compared to Prohibition, one of the stupider ideas the US ever came up with, and clearly the STUPIDEST thing ever put into the US Constitution.
IMHO Mysterium was somewhere on par with Chronoliths, but The Harvest (Have you read it?) was much better. Haven't seen Memory Wire or A Bridge of Years. Mysterium started with a really neat premise and followed up on it, but then seemed to me to drift along after that. Perhaps I would have enjoyed it more if I knew was gnosticism was, I'll admit.
No, that was "Back to the Future II"
on
The Chronoliths
·
· Score: 2
Slashdot Mirror
I wasn't aware they still were making that stuff. It would also be neat to see some original work in larger technologies, specifically tailored to the super-reliability market. I guess it would be too expensive. Your remark about 'lots of atoms to knock out of several microns' is exactly on the mark, whether the knocking is being done by radiation, phonons, or electrons of normal operation. But in today's technologies, just about every structure is submicron.
"Press Enter" by John Varley
or
"The Adolescence of P One" by
for tales of AI gone bad. There are others...
Human: "Is there a God?"
zzzaaaappppp - lightning strike fuses the power switch on.
Computer: "Now there is."
I agree with you 100%, and have long felt that the NRA is the right example. More appropriate than many might thing, even. After all, crypto is a big geek issue, so we are very much into, "The right to bear arms." But getting too close to the NRA might make for strange bedfellows. (except for ESR, of course)
People like Hollings for non-re-election? Perhaps we need "The Geek Lobby Page" where information about key publicly elected officials is kept.
When is Hollings up for re-election?
Who is running against him?
Are the opponents views any better?
We all grumble, complain, and flame. We also say we're too small. But have we tried yet to use tried-and-true mainstream political techniques?
I agree with everything you say.
But I believe that long-term space operation is just a place where current electronics haven't ever gone before. Current electronics are obsolete and in the landfill well before hitting the wearout limits. In truth, I believe relatively little has been done about investigating true wearout compared to the quality control issues you mention.
But IMHO current electronics simply won't last the way older, larger geometries did. OTOH, if we were to take today's quality capabilities and knowledge, and build some appropriately larger geometries, we truly could build long-duration space electronics.
I seriously question the long-term of any semiconductor electronics built today. No, there are no moving parts - except the electrons and any atoms they may knock about as they scurry on their way from source to drain and through the wires.
Shipping reliable semiconductors has always been a lifetime issue. There is a "bathtub curve" of failures, with a higher number of early fallout, then a very reliable main lifetime, then failures rise again at wearout. Wearout happens through mechanisms like electromigration, where the electrons physically knock the metalization atoms out of place. In addition, all of the hot process steps like diffusion continue to happen, just at much slower rates. High reliability semiconductors are "burned in", run at higher temperatures and voltages than normal, to force them past that early fallout and throw those parts away.
So what does this mean to space electronics? First, radiation just doesn't help. You can design rad-hard, but the crystal lattice is still taking damage, and it's cumulative. The low temperature helps to slow down wearout mechanisms.
But the big problem is modern technology. The smaller geometries will simply wear out faster. Finer wires are more subject to electromigration, though using copper is an improvement because the atoms are heavier than aluminum. But gates are thinner, as are diffusions and spacings, non of which helps long life. When designing a burn-in regimen, it's getting tougher to get past early failures without approaching wearout. While frequency can be reduced to increase lifetime, scaling voltage down is getting tougher, because we're running darned close to minimums, already.
One of my pet thoughts is the idea of electronics for a multi-generation starship. Other than slowing it down, stopping as much as possible, reducing voltages, etc, it's a tough problem. Maybe the best way is to scrape the bargain bins for old technology.
Nor am I terribly serious about photography, but my family and I do like our vacation slides. We used to use the Nikon FA for slides, and the FT2 for snapshots, but the latter has been waiting for a round tuit for too many years to get the light meter fixed. So the FA has been swapping between print and slide film, unconvenient at times.
In the medium run, I see moving the FA back to slides, and getting a digital for snapshots. That leaves me with a serious controllable camera when it really matters, and pre-selection on snapshots when that matters.
In the longer term, I don't know what we'll do. I have a small collection of Nikon mount lenses, though I guess they're all obsolete by any contemporary terms. Certainly predates autofocus. I wish someone would come up with a digital SLR back too, and upgradable CCDs would sure be a good idea, while they're at it.
But perhaps the SLR/prism viewer itself is obsolete, given that the CCD can deliver a preview image just fine. But the whole eyepiece viewing and control is certainly more precise than a mini-preview screen. Perhaps if they could deliver the image through an eyepiece and keep the better muscular control.
Case Western Reverse University,
at least when I went there. aka See double, you are you.
Are the Spitwad and the Fountain still in the quad? Is Presti's Donuts still in business, and are the Best Cinnamon Rolls in the known universe still available there at 1:15 AM?
Doesn't fit But I Play One On TV. (BIPOOTV) Kindly de-acronymify?
In that case, read "Sundiver" by David Brin. Good book, first of the first Uplift Trilogy. (Though IMHO the second book, "Startide Rising" was better, and the third, "The Uplift War" was somewhat weak.
IANAL, but...
This whole thing seems familiar, not just with GIF but with Rambus and the SDRAM/DDR standards.
At the time, one of the writings mentioned a thing called "Equitable Estoppel." My interpretation was that if you had a patent that was becoming an industry standard, you had to begin notification "promptly," and to allow it to become a standard and *then* begin notifying/litigating was legally naughty.
Rambus is still around, though a shadow of their former arrogance. (I understand that the people are still just as arrogant as ever, they just don't get the press.) In some ways, notably submarining and patent-stretching Rambus was worse. But at least once they had stretched their original art to look like it covered SDRAM and got it issued, they were prompt in filing suit.
It looks like this company deserves no less.
Doesn't say in the article, haven't had time to wade through the PDF. Other comments have mentioned opt-out, but nothing concrete on how to do so.
Simple question, whoever gives a good, simple reply deserves to be mod'ed up.
>I used to really believe in total free-market capitalism
I suspect you and a lot of others are missing the point here, though you've eventually come to the right conclusion.
People are not capitalists, the collective is. Capitalism is not based on people believing in it, it's based on people being greedy. Government oversight is *necessary at the theoretical level* for capitalism because the participants need not believe in it, and the most successful participants generally don't. Pretty much every successful business' goal is to corner the market and sit back collecting money. No business has as its long-term goal to engage in continuous cutthroat competition with its peers. No business has as its long-term goal to reduce prices and improve value for the customers. Those are not long-term goals, those are necessary means to the real long-term goal - making money.
Nothing wrong with this. Capitalism simply recognizes that greed is an effective motivator. But it requires recognition that an outside regulating force is necessary in order to prevent the end Marx predicted.
Unfortunately in the past decade or so, people have come to worship the free market without realizing its inherent instability, and have worked to remove the external corrective forces.
Others have mentioned Microsoft's giant cash pile, and said that they can keep giving away Windows practically forever in the third world.
But it doesn't matter, this HURTS.
First off, there's the simple fact that Linux made it onto the table, and it took Bill Gates himself to go 'bring back the account.' That's quite an endorsement Microsoft has given Linux, right there.
Second, Linux made it onto the table, here and in Norway. Simply getting onto the table is new. Maybe for the next few rounds Linux will get onto the table as a bargaining ploy. But one of these times it's going to be a WIN.
Third, Microsoft is sitting on a big pile of cash, just about everyone knows that. But now a big part of their invincible image has become that pile of cash. It's no longer completely fluid, it has become an essential holding. Besides, the death of a thousand paper cuts is perhaps more likely than a big wound. So add up X-Box subsidies, third-world givaways, European givaways, (Norway, anyone?) and no doubt some 'License V6 adjustments' to keep customers in the fold, and it's going to be tough to keep that wad of cash.
That cash is also only part of the image. The Gates trip to Peru chips away, License V6 chips away, the DOJ trial took a big chunk out, etc.
I can't wait for hordes of Elvis fans to protest against the title of, "Return of the King," or will the be in the theatres apoplectic when they find that the king who returned *wasn't* Elvis.
Or since there have been any number of treatises examining Lord of the Rings for Christian elements and Nuclear warnings, etc... How about evidence that JRRT could see the future and named the book accordingly?
Methinks you need a humor-o-meter, here.
Otherwise, Ms. Portman, naked and petrified, played bit parts in numerous scenes. Most notable was her 'falling rock' part in the stair scene during the Balrog attack. Some of her other scenes were largely covered by vegetation, so I guess they would only qualify as petrified.
Actually there are secure extensions to both telnet and smtp. The security problem can be separated from the protocol itself, and there are several such implementations. By the same token, with ssh and ssl/tls the same exists. Agreed it's a little tougher to see, once the bytes get encrypted, but the scheme is carefully documented and can be traced *by the authorized parties*.
Actually, FWIU there *is* something wrong with .doc. Again, this is all gleanings, but to sally forth into the darkness:
.doc has none of those three attributes. If it did, someone would have created really good .doc import/export filters by now, not just almost-decent ones. There are some stories about that .doc is not even documented inside Microsoft, rather that there is a 'reference implementation' of source code.
.doc format appears to be engineered largely to force users to upgrade.
.doc.
"Good" protocols are things like telnet, smtp, etc. They are simple, straightforward, and discoverable. It seems that
A "good" file format, from a technical perspective, would have offered much better compatibility between revisions. The
I'd say that there are some intrinsic problems with
The current administration has not been very friendly to such highly regulated non-business-friendly things as a public IP network. The concept of 'public good' seems to have been tossed out the window. OTOH, the current round of business scandals certainly sets the tone for such a thing. Several of the current scandals threaten to darken serious portions of the Internet, at least temporarily. My own home connection (Adelphia) is threatened, for that matter.
On the more historic side, apparently nobody remembers The Source or any of the early highly-proprietary online services. Even CompuServe, AOL, and Prodigy only survived by becoming Internet portals, though all but AOL are all but gone. People simply didn't want to be locked in. They wanted the "IP Utility" that the Internet originally offered. Ever since the Internet was privatized, there's been a tug-of-war to turn it back into a proprietary cash-cow, despite the teachings of even recent history.
But then again, we went to the Moon, and threw all of that away.
She's 79 and has difficulty from time to time, but it really helps that I can ssh in.
Yeah, Hiro did have Reason, but the sword was in the Metaverse, not reality, and he did hack his way through at least some of the problems. Uncle Enzo is allowed to use more physical things, because that's where his power was rooted - he wasn't a hacker.
But this one was a bit more fisticufish than some other Stephenson. I was thinking more of Cryptonomicon or Diamond Age. A little of the hero doing fisticuffs there too, but still reasonable.
I seem to remember hearing that some of the German rocketry work was done by Jewish slave labor under inhuman conditions. Don't know if it was as bad as the medical experimentation, though.
IMHO, the secondary damage of the War On Drugs far outweight the damage of the drugs, themselves. By trying to interdict drugs, we turn them into a high-dollar business. Also IMHO, it becomes self-defeating, because the more successful you are at interdiction, the more the cost rises, the better-financed the drug supply chain become.
I heard on NPR that in the 1968 election, Nixon promised to reduce crime. After getting elected, he had to produce. Some of his advisors told him that all of the more traditional law enforcement techniques had been proven not to work. (by experience) His best shot was drug treatment, to reduce demand, and therefore crimes of financing. He went along with it, it worked, and crime actually did go down measurably. In the 1972 election crime was no longer a big issue, so he dismantled the apparatus, and the approach has never been taken seriously again.
Kind of like the way Clinton/Greenspan actually did achieve a "soft landing" of the economy right before the dot-com boom. But now the concept appears to be forgotten, and I have no doubt that whenever recovery comes, it will be back into the usual boom/bust cycles.
Back to topic, filesharing is an interesting comparison to drugs because it is a widespread crime. Perhaps it should be better compared to Prohibition, one of the stupider ideas the US ever came up with, and clearly the STUPIDEST thing ever put into the US Constitution.
IMHO Mysterium was somewhere on par with Chronoliths, but The Harvest (Have you read it?) was much better. Haven't seen Memory Wire or A Bridge of Years. Mysterium started with a really neat premise and followed up on it, but then seemed to me to drift along after that. Perhaps I would have enjoyed it more if I knew was gnosticism was, I'll admit.
1995: Florida Wins World Series