The problem is, NASA has restricted manned space flight in the United States to itself, and done their best to make sure that such flight remains as expensive and unavailable as possible.
Currently they have a monopoly by law in the USA on supplying human-to-space transport, and I don't see anything in the article which indicates they are ready to hand over such authority to anyone else. Their idea of "privatization" has tended to be "We'll sell you a few kilograms of mass and a couple cubic feet to put a machine in, and have an astronaut push a button to start and stop it for you". Maybe they're thinking about allowing nominal control over parts of the space station where passengers they transport can work.
If making space transport cheap and common was their focus, the DC-X, based mainly on proven engineering, would be in full development instead of the "technology demonstrator" projects that got the money instead. NASA wants to need a zillion personnel for each launch, not to mostly disband while reliable automatic systems and a few air traffic controllers take over. D'oh!
I sincerely doubt that a colony, if successful, would represent radical change. The first european colonies in the New World tried radical ideas such as communal ownership and responsibility for labor, and, as was inevitable, many starved to death in the first winters. The same would occur in any off-earth colony that made the same errors and tried to operate without a hierarchy of some sort (I would hope democratic), forbade ownership of private resources, and did not use authority to control and direct the labor and activities. Only quicker, as the consequences would include a lack of air to breathe as well as food to eat.
I'll have to plead no contest on this. On my own systems I always use ECC memory to avoid the possibility of a stray cosmic ray or alpha particle scrogging a program. This of course begs the question of whether ECC RDRAM is even available or usable by the i820, I haven't investigated past finding the i820 conspicuously absent from Intel's chipset comparison page. At some points in time, particularly during the time of the "logical parity" scam (a cheap parity generator chip on SIMMS that generated signals making it appear there was real parity memory present), it was almost impossible to get ECC-capable EDO SIMMS from any source for reasonable prices.
But on the other hand, the probability of power fluctuations or ordinary software bugs destroying data is probably a couple of orders of magnitude greater than RAM being at fault. In my experience, particularly if Microsoft software is being used, system failures tend to be so common that real memory problems almost aren't worth investigating, or are impossible to detect due to being concealed down in the "noise". I've been told "Uh, our memory is only guaranteed to work with Microsoft products" when trying to return defective RAM that generated repeatable Signal 11 errors compiling Linux kernels. The "flaky" merchants involved depending on blaming Microsoft software for their own defective products.
The only thing close to a solution is to use a good UPS, "backup, backup, backup", take computed results with a grain of salt, and retain those backups for a long time. Admittedly this won't fix the "game going zing" problem, but if this is important enough to you I suggest saving the game's state regularly. Provided the game lets you do so.
All life is a risk, when you come right down to it.
To release a single ferite core to fall out, not only would both the X and Y address lines which were pulsed to read/write the core be broken, disabling a large number of bits, but also the sense wire which was threaded in a serpentine manner through all the cores in a memory bank, rendering the entire bank unusable.
Spare banks may have been put in to account for defective cores or wiring errors, but I've never heard of a core memory system with the ability to switch to new banks in the field without the help of a repair technician.
This kind of thing is typical when rolling out entirely new technologies like RDRAM, although with the i820's relatively long development time, and the delays already experienced, Intel should have been expected to have already found and corrected the problem.
No one in their right mind would be deploying RDRAM or i820-based boards in server or other mission-critical applications at this early stage of development anyway - it simply hasn't been proven to work reliably yet, and given what little benchmarking I've seen comparing performance with AMD's and existing Intel chipsets, for home and gaming use the supposed performance improvements seem more a promise for the future than a current reality, rather like AGP before AGP 2X was implemented. With the caveat that this could, of course, be due to the use of preliminary or experimental BIOS versions in the machines being benchmarked.
Advice: wait until it has been proven to work, before jumping onto the Rambus.
... is that he isn't stable enough to run a decent magazine column, or capable of handling the heavy load represented by evaluating entire operating systems against each other.
But does anyone but me remember when, shortly after Zilog introduced their Z80 microprocessor, they sued a company called "Z-Systems" under the theory that their Z80 trademark gave them the rights to the letter "Z"?
Well, the 8088 was an incremental improvement on the 8080/8085, done in, as I recall, about a week or so when it became clear Intel's hugely expensive iAPX432 project was going to be a tremendous flop. Everything since up to the PIII has been more incremental improvements.
So, Merced will actually be starting the third actually new microprocessor family Intel has ever made, the second one having died a-borning.
Reference the case Bozo vs. Bozo, won by the "Bozo's Resturant" owner who had apparently been using the name long before Larry Harmon. Mr. Harmon was reportedly livid at the result, expressing his anger and disbelief that the small resturant owner was allowed to continue using the name "Bozo" for that mere reason, as opposed to him being awarded exclusive use of the name based on the size of his great "Bozo the Clown" empire.
It seems to have become the vernacular for the monitoring techniques as well. In any case you can download "anti-Tempest fonts" to make the monitoring harder from the site linked to in the previous comment.
Netscape 4.61, on a Windows 98 box, ploinks up the "Do you want to buy Quicktime now or keep on being pestered each day until the end of time" box, and displays a test PNG in the center of the window.
MS Internet Explorer just puts up in the center of the screen. Treats a PNG rather differently than a GIF or JPG file, anyway.
However! Netscape 4.51 on a Red Hat 6.0 Linux box displays the PNG image exactly as it does an identical image compressed into a GIF.
If websites were forced to switch to this format that is evidently only supported by a major web browser on that OS (by my quick check), I applaud the positive effect for Linux. But is the world ready to have every websurfer a sysadmin?
I spent a frustrating two years or so trying to help a company I won't name into being an Internet Service Provider, with explicit permission and encouragement to do so. Unfortunately, the president of the company decided that he was going to be the sole and only salesman for the services; he knew little or nothing about the Internet and displayed no willingness or ability to learn about it, preferring to operate on the assumption that whatever he wished to be true about the business would magically become the case, and blaming any failure of reality to mold itself that way on his subordinates.
The result was the predictable disaster. He was told that because not all customers would be using the proposed 250 dialin connections at once, statistically the 250 connections could support perhaps as many as 2000 users. This caused him to jump to the conclusion that all 2000 could dial in at the same time, 8 customers using each a modem and ordinary phone line simultaenously, and therefore only 32 modems would be needed to support 2000 customers by the statistical rule. Then I was presented with 4 analog dialin lines and modems, and a statement that they should be able to support 250 users, which we would start the business off with.
Upon reality being presented, along with the original provisioning documents, the idea of selling dialin lines was suddenly dropped, and reselling T1 service from our connection was declared to be our new Internet business; also it was proposed that we would be selling web sites, though already web farms with much higher bandwidth were coming into use.
Then everything seemed to - oscillate. One week, it would be declared that we were going to put websites up, a couple of weeks later we were going to go full bore selling bandwidth. No details were ever available of the "plans".
Eventually this climaxed in an angry, out-of-the-blue mobile phone call demanding "Yes or no - do we have a clear T1 channel to the Internet?" and escalating fury as I tried to explain that that was a trick question; we had a T1 to a backbone ISP, but no guarantee there was 1.544mbps to each and every server on the net, et cetra. Upon being told I would be fired if I did not simply answer the question with a "Yes" or "No", I said "Yes", true in the sense we had a "clear channel" DS1 line and not a Frame Relay connection. Nothing more happened for a while, then I was called into the president's office and was told he had sold some T1 connections - and that he decided, and had told the companies involved, that they did not have to actually purchase a point-to-point connection to us, or any CSU/DSU or router; because we already had the T1 & et cetra, we would just sell them "part of ours". Magically wishing the bits from our networks into the customers'. The response he was given, which as it had to be was that the president/salesman had been given specific lists of the required equipment down to part numbers and itimized costs, was not well accepted - I was told, in effect, that he was too important a person to be required to read such trivia as what subordinates put into his "In" box.
Probably I should have known things were not going to work out well when the principal involved proposed early on that we could save much money by buying an AOL account or two to resell to our high-bandwidth customers. Sigh. At least they finally learned to use email, but it was rather a relief to leave that place.
We US citizens with Canadian relatives get special perks on hearing the best Newfie jokes, BTW.:-)
Re:"A US Army official was not available for comme
on
U.S. Army Testing Jini
·
· Score: 1
I've noticed. For that matter, I approve.:-)
"A US Army official was not available for comment"
on
U.S. Army Testing Jini
·
· Score: 1
I can't help pondering if this meant the US Army is being deliberately closed-mouth, or if the Reuters reporter was was merely stating the fact that whichever one "official" he picked at random happened to be in the latrine when called...
The cheetahs were doomed, all agreed; in fact, they had been doomed for millons of years, as far as anyone could tell. Any day now... well, any eon now...:-)
Item: aquatic blackfly larvae in northern climates often make up the majority of food that trout fry and other fish consume at certain ages.
Item: It is becoming increasingly popular to dose bodies of water heavily with a bacterium which kills blackfly larvae along with other insects, to prevent tourists - often mainly sport fishermen - from being annoyed by these incredibly irritating flies.
Can anyone besides me see a stunningly stupid mistake being made?
Has any species anywhere ever become extinct because of a loss of "genetic diversity"? The usual result (walruses, cheetahs) seems to be the development of species superbly adapted to their roles, and disconcertingly robust despite predictions of disaster by geneticists.
The wolf population on Michigan's Isle Royale illustrates the principle quite well. Since a single pack colonized the island in the 1950's, each major drop in the population seems to have caused the experts studying them to pronounce them doomed.
The wolves (currently on the increase), apparently refuse to believe the predictions.
Nature will take care of it. If 80% of the birds are unfit, they will be eaten, fly into tree trunks, etc, and be quickly replaced by the most competent of the remaining 20% and their offspring.
Slashdot Mirror
The problem is, NASA has restricted manned space flight in the United States to itself, and done their best to make sure that such flight remains as expensive and unavailable as possible.
Currently they have a monopoly by law in the USA on supplying human-to-space transport, and I don't see anything in the article which indicates they are ready to hand over such authority to anyone else. Their idea of "privatization" has tended to be "We'll sell you a few kilograms of mass and a couple cubic feet to put a machine in, and have an astronaut push a button to start and stop it for you". Maybe they're thinking about allowing nominal control over parts of the space station where passengers they transport can work.
If making space transport cheap and common was their focus, the DC-X, based mainly on proven engineering, would be in full development instead of the "technology demonstrator" projects that got the money instead. NASA wants to need a zillion personnel for each launch, not to mostly disband while reliable automatic systems and a few air traffic controllers take over. D'oh!
I sincerely doubt that a colony, if successful, would represent radical change. The first european colonies in the New World tried radical ideas such as communal ownership and responsibility for labor, and, as was inevitable, many starved to death in the first winters. The same would occur in any off-earth colony that made the same errors and tried to operate without a hierarchy of some sort (I would hope democratic), forbade ownership of private resources, and did not use authority to control and direct the labor and activities. Only quicker, as the consequences would include a lack of air to breathe as well as food to eat.
Only if you're good enough to keep them from ever finding out.
I'll have to plead no contest on this. On my own systems I always use ECC memory to avoid the possibility of a stray cosmic ray or alpha particle scrogging a program. This of course begs the question of whether ECC RDRAM is even available or usable by the i820, I haven't investigated past finding the i820 conspicuously absent from Intel's chipset comparison page. At some points in time, particularly during the time of the "logical parity" scam (a cheap parity generator chip on SIMMS that generated signals making it appear there was real parity memory present), it was almost impossible to get ECC-capable EDO SIMMS from any source for reasonable prices.
But on the other hand, the probability of power fluctuations or ordinary software bugs destroying data is probably a couple of orders of magnitude greater than RAM being at fault. In my experience, particularly if Microsoft software is being used, system failures tend to be so common that real memory problems almost aren't worth investigating, or are impossible to detect due to being concealed down in the "noise". I've been told "Uh, our memory is only guaranteed to work with Microsoft products" when trying to return defective RAM that generated repeatable Signal 11 errors compiling Linux kernels. The "flaky" merchants involved depending on blaming Microsoft software for their own defective products.
The only thing close to a solution is to use a good UPS, "backup, backup, backup", take computed results with a grain of salt, and retain those backups for a long time. Admittedly this won't fix the "game going zing" problem, but if this is important enough to you I suggest saving the game's state regularly. Provided the game lets you do so.
All life is a risk, when you come right down to it.
This seems a topologically unlikely story.
To release a single ferite core to fall out, not only would both the X and Y address lines which were pulsed to read/write the core be broken, disabling a large number of bits, but also the sense wire which was threaded in a serpentine manner through all the cores in a memory bank, rendering the entire bank unusable.
Spare banks may have been put in to account for defective cores or wiring errors, but I've never heard of a core memory system with the ability to switch to new banks in the field without the help of a repair technician.
This kind of thing is typical when rolling out entirely new technologies like RDRAM, although with the i820's relatively long development time, and the delays already experienced, Intel should have been expected to have already found and corrected the problem.
No one in their right mind would be deploying RDRAM or i820-based boards in server or other mission-critical applications at this early stage of development anyway - it simply hasn't been proven to work reliably yet, and given what little benchmarking I've seen comparing performance with AMD's and existing Intel chipsets, for home and gaming use the supposed performance improvements seem more a promise for the future than a current reality, rather like AGP before AGP 2X was implemented. With the caveat that this could, of course, be due to the use of preliminary or experimental BIOS versions in the machines being benchmarked.
Advice: wait until it has been proven to work, before jumping onto the Rambus.
... is that he isn't stable enough to run a decent magazine column, or capable of handling the heavy load represented by evaluating entire operating systems against each other.
Funny, yes!
But does anyone but me remember when, shortly after Zilog introduced their Z80 microprocessor, they sued a company called "Z-Systems" under the theory that their Z80 trademark gave them the rights to the letter "Z"?
I kind of liked the undocumented instructions, like the ones that let you use the upper & lower bytes of the IX and IY registers as H and L.
Sped some code up quite a bit, actually, and the undocumented instructions persisted through the faster Z80 versions.
Well, the 8088 was an incremental improvement on the 8080/8085, done in, as I recall, about a week or so when it became clear Intel's hugely expensive iAPX432 project was going to be a tremendous flop. Everything since up to the PIII has been more incremental improvements.
So, Merced will actually be starting the third actually new microprocessor family Intel has ever made, the second one having died a-borning.
Reference the case Bozo vs. Bozo, won by the "Bozo's Resturant" owner who had apparently been using the name long before Larry Harmon. Mr. Harmon was reportedly livid at the result, expressing his anger and disbelief that the small resturant owner was allowed to continue using the name "Bozo" for that mere reason, as opposed to him being awarded exclusive use of the name based on the size of his great "Bozo the Clown" empire.
It seems to have become the vernacular for the monitoring techniques as well. In any case you can download "anti-Tempest fonts" to make the monitoring harder from the site linked to in the previous comment.
Emperically testing:
Netscape 4.61, on a Windows 98 box, ploinks up the "Do you want to buy Quicktime now or keep on being pestered each day until the end of time" box, and displays a test PNG in the center of the window.
MS Internet Explorer just puts up in the center of the screen. Treats a PNG rather differently than a GIF or JPG file, anyway.
However! Netscape 4.51 on a Red Hat 6.0 Linux box displays the PNG image exactly as it does an identical image compressed into a GIF.
If websites were forced to switch to this format that is evidently only supported by a major web browser on that OS (by my quick check), I applaud the positive effect for Linux. But is the world ready to have every websurfer a sysadmin?
Translation requests are now coming back untranslated.
I spent a frustrating two years or so trying to help a company I won't name into being an Internet Service Provider, with explicit permission and encouragement to do so. Unfortunately, the president of the company decided that he was going to be the sole and only salesman for the services; he knew little or nothing about the Internet and displayed no willingness or ability to learn about it, preferring to operate on the assumption that whatever he wished to be true about the business would magically become the case, and blaming any failure of reality to mold itself that way on his subordinates.
The result was the predictable disaster. He was told that because not all customers would be using the proposed 250 dialin connections at once, statistically the 250 connections could support perhaps as many as 2000 users. This caused him to jump to the conclusion that all 2000 could dial in at the same time, 8 customers using each a modem and ordinary phone line simultaenously, and therefore only 32 modems would be needed to support 2000 customers by the statistical rule. Then I was presented with 4 analog dialin lines and modems, and a statement that they should be able to support 250 users, which we would start the business off with.
Upon reality being presented, along with the original provisioning documents, the idea of selling dialin lines was suddenly dropped, and reselling T1 service from our connection was declared to be our new Internet business; also it was proposed that we would be selling web sites, though already web farms with much higher bandwidth were coming into use.
Then everything seemed to - oscillate. One week, it would be declared that we were going to put websites up, a couple of weeks later we were going to go full bore selling bandwidth. No details were ever available of the "plans".
Eventually this climaxed in an angry, out-of-the-blue mobile phone call demanding "Yes or no - do we have a clear T1 channel to the Internet?" and escalating fury as I tried to explain that that was a trick question; we had a T1 to a backbone ISP, but no guarantee there was 1.544mbps to each and every server on the net, et cetra. Upon being told I would be fired if I did not simply answer the question with a "Yes" or "No", I said "Yes", true in the sense we had a "clear channel" DS1 line and not a Frame Relay connection. Nothing more happened for a while, then I was called into the president's office and was told he had sold some T1 connections - and that he decided, and had told the companies involved, that they did not have to actually purchase a point-to-point connection to us, or any CSU/DSU or router; because we already had the T1 & et cetra, we would just sell them "part of ours". Magically wishing the bits from our networks into the customers'. The response he was given, which as it had to be was that the president/salesman had been given specific lists of the required equipment down to part numbers and itimized costs, was not well accepted - I was told, in effect, that he was too important a person to be required to read such trivia as what subordinates put into his "In" box.
Probably I should have known things were not going to work out well when the principal involved proposed early on that we could save much money by buying an AOL account or two to resell to our high-bandwidth customers. Sigh. At least they finally learned to use email, but it was rather a relief to leave that place.
We US citizens with Canadian relatives get special perks on hearing the best Newfie jokes, BTW. :-)
I've noticed. For that matter, I approve. :-)
I can't help pondering if this meant the US Army is being deliberately closed-mouth, or if the Reuters reporter was was merely stating the fact that whichever one "official" he picked at random happened to be in the latrine when called...
The first I read of "instant Messages" was in Cordwainer Smith's Norstrilia novel, a combination of two stories published in the early 1950's.
Worse! The things are grown using radiation from the largest nuclear reactor in the solar system! Eieeeee!!!
The cheetahs were doomed, all agreed; in fact, they had been doomed for millons of years, as far as anyone could tell. Any day now... well, any eon now... :-)
Item: aquatic blackfly larvae in northern climates often make up the majority of food that trout fry and other fish consume at certain ages.
Item: It is becoming increasingly popular to dose bodies of water heavily with a bacterium which kills blackfly larvae along with other insects, to prevent tourists - often mainly sport fishermen - from being annoyed by these incredibly irritating flies.
Can anyone besides me see a stunningly stupid mistake being made?
Has any species anywhere ever become extinct because of a loss of "genetic diversity"? The usual result (walruses, cheetahs) seems to be the development of species superbly adapted to their roles, and disconcertingly robust despite predictions of disaster by geneticists.
The wolf population on Michigan's Isle Royale illustrates the principle quite well. Since a single pack colonized the island in the 1950's, each major drop in the population seems to have caused the experts studying them to pronounce them doomed.
The wolves (currently on the increase), apparently refuse to believe the predictions.
Nature will take care of it. If 80% of the birds are unfit, they will be eaten, fly into tree trunks, etc, and be quickly replaced by the most competent of the remaining 20% and their offspring.
:-)
This has worked well in the past.
Damn, I forgot the dynamic address allocations for the virtual particles in the quantum foam...