I simply write all my docs with inverted reality. If my research has shown that the speed of object x is fourteen furlongs per fortnight, I document that the speed is NOT fourteen fpf. Bwaaaaahahahaaa !
What I want to see in a filesystem is a combination of Logical Volume Manager, Journaling and Encryption. This would provide quick booting even with large mounted partitions ( a feature of journaling ) the ability to have a volume larger than a single physical disk drive ( a feature of LVM ) and the security of having the data encrypted. I would think that by integrating these features into a single filesystem there may be some benefit in performance over trying to glue seperate implementations of these features together. There, I did the hard part - creating the idea. Now one of you hacker types, please begin coding !
In a perfect world, a.0 release of any software could be trusted to be correct. So far, my world is far from perfect. As a sysadmin running production on linux boxen, I would never consider putting a.0 release of ANYONES distribution on a production machine. At the very least, I would wait a few weeks till all the most obvious flaws had been addressed with patches etc. If I were force ^h^h^h^h^h strongly encouraged to use a.0 release somewhere, I would sure as hell put it on a test box as long as possible before moving it onto production. I will not say either good things or bad things about RedHat or any other vendor. As far as I am concerned, like the IRS, they just are. I deal with their existance and try to keep my life easy and stable - I have found ways to use their efforts and products to my advantage and so far have not been burned too bad. When I do have problems, with an honest apprasal of the situation I realize that there was something I could have done ( some might even say should have done ) to avoid the problem in the first place.
I think an important quality of Internet access which is desirable for the evolution of society into cyberspace ( opening myself up a bit there eh? ) is that Internet access must be considered as always on, no cost and fast. It needs to be so cheap and convienient to use that there is simply no thought of not using it for whatever communication / research / play / whatever task is at hand. Ie do not want to continue or exacerbate the class dichotomy of the computer/internet haves vs. the have-nots. By charging per unit access, poor people will have less access to Internet, which will serve to keep them poor. Sounds like bad juju to me.
Datalens seems to be down. Could be something as simple as a telephone guy stealing a pair from their uplink T-1. This can take days to fix, since the telco will say, " no, we havent done anything to YOUR circuit ". I have seen T-1 lines go away for 3 or 4 days from causes like this. How long has the site been down? Does anyone know how big of an ISP Datalens is ?
Will this release permit the development of everyman encryption devices ? One of the common rumors (facts?) of our time is the government intercepting all communication and computer-analysing it for evil words etc. Will someone now make nifty jeez-oh pocket encryptors cheap and easy enough for everyman to use - something to go between a telephone set and the public network, or perhaps between a microphone / speaker and a radio set ? Is the lack of these products in the real world (TM) proof of excessive government controls and insidious dark forces ? And what about Naomi ?
San Ramon Unified School District in conjunction with Rotary Club sponsored a web programming contest. This contest was for teams of students who would form businesses providing web site design for business customers, who were participating in the contest. The teams were evaluated on both the technical merit of the website and their ability to communicate well with the "client" and extract meaningful knowlege of the business so as to be able to build an effective web site. This Link provides more information on previous contests.
I wonder about the value of overclocking and pushing technology to the extremes in home computers, when the designers of supercomputers have realized that the performance gains achieved by pushing clock speed do not scale linearly with the cost. At some point a maximum speed will be reached and the cost of going any further will be staggering. The architects of high end systems have realized that massively paralell systems were the most practical way to obtain orders of magnatude increases in performance. I think that effort designing languages and programming pardigms to effectively use MPP architecture may be effort better spent than building fluorinert or alcholhol cooled overclocked systems. I suppose these systems are more akin to chopped channeled and supercharged 43 ford roadsters than to a 747 in terms of transporting people and cargo.
Consider an intentionally designed and created virus. And biologically active viruses do go from species to species - most influenza in china comes from birds, particularly chickens.
DNA based computers might be much more sensitive to the effects of radiation, or other mutagenic factors. I can just see the label on the computer, " Caution, the hacker general has determined that smoking causes crashes ".
One interesting tool used in DNA research is called PCR, which is a way of multiplying the amount of DNA in a sample. 24 cycles of PCR amplification would yield a 2 ^ 24 increase in the quantity of a DNA based circuit, and takes less than a day to do. For once, memory could grow faster than data!
I would be concerned about the possibility of your computer infecting you with a biologically active virus....
We have several VA boxes here, and paid extra for "enhanced support" ( one of the joys of working for a very large multi-national corporation - they buy enhanced support for EVERYTHING ). When we had problems using VA supplied gigabit ethernet cards, VA sent two people out here for a couple of days to help us and we did eventually fix the problem.
My perception of VA's linux system support is that they provide a reasonably recent distribution with additional drivers for any uncommon hardware that they provide ( we have a Mylex extreme raid controller ). I think ( and have suggested to them ) that they should provide regular system updates - sending out new CDs with the latest kernels and patches. They could send the cds out quarterly perhaps even with a special update when a new version of whatever distribution ( we got redhat from them ) is released. A newsletter would also be a good idea. They need to go beyond having just good solid products and work on evangalisim, and going the extra mile in customer service is one way to do it. VA systems do cost more than some other companies for the same motherboard / cpu speed / memory configuration. I like their chassis better than the cheaper ones ( Boombox for example has very poor cooling design ). changing from "yes we have customer service " to " customer service is our most important product " would help VA. Many of the things they do - for example employing and supporting the development projects of some of the leaders in linux GUI software, as well as other significant contributors to linux are good things for a company to do. They help the whole linux community, but are probably not recognized by the majority of VA's customers. Having that extra mile customer service would be recognized.
Subject number is customer service for Etoys. I am an E-toys customer ( up untill today ). I called them and the person I spoke with was interested and reasonable and said I would get a call back. Perhaps if more people called, they may realize the scope of discontent with their actions. ( at $.10 per minute LD 800 charges )
Perhaps we should all wear togas ? At any rate, this should be a blast. Rather than the hit or miss chance meetings of fellow posters around a/. booth, this may be a good opportunity to meet others of this here fine community ! !
Hmmm. I wonder what they are going to open beyond what has always been available ? The only things that seem to be hidden are related to domain owner authentication and billing. This may make it easier ( or perhaps only different ) for spam generators to get the contact info for large numbers of domains, but to my knowlege the info has always been there. In fact, I remember getting spammed by someone selling a list of domain managers a while back.
I think the most important aspect of this new chip is the continued pressure of competition on Intel. This should increase the motivation for the continuation of gains in performance and cost / performance ratio. In terms of what Joe Doakes average hobbiest gets for the money, I think the optimum design right now is a dual processor system using a not-quite bleeding edge CPU - say a PII 450 or therabouts. The cost of adding a second CPU to a system seems to be less than the cost of bumping up the clock another hundred or so MHZ. This does not account for improvements in internal efficiency, but you do get a lot more bang for the buck with 2 CPU's. The more people who have multiprocessor systems, wether SMP, clusters or even shared-memory systems the faster we will see the evolution of new software paradigms that will truly advance the state of the art. Z
I worked with a client who had 2 IR Wireless links (LACE - Laser Atmospheric Comm. Equipment ? ) installed. The systems were configured for T-1, although they could be used for Ethernet at 10Mbps. The path length was about 1/2 mile in each case. One link was pretty reliable, the other was quite troublesome. In the troublesome link, one end was mounted atop a chimney ( can you say wind? Sure! ) and the other was on a floating membrane roof. The links cost about $37K each, including the T-1 to Fiber converter and the mounting hardware. They were inconvienient to work with due to the mounting locations and hardware, and when aligning them, very minute adjustments of knurled-headed screws made significant changes in signal. The complete range of usefull adjustment was on the order of one rotation of the screw!
I imagine that there are applications where these types of things are usefull, but the installation has to be rock-solid and well engineered.
I think a better growth path is to install fiber. Lots of fiber. Everywhere. NOW !
I, for one would like to see a set of benchmark results for Linux that would help a person to make decisions about the configuration of hardware platforms for systems. I would like to see a set of test results for different types of system activity, such as compiling code, raytracing/graphics/visualization, file system access, network bandwidth, combined network/filesystem access etc. This set of measurements could then be run on a variety of hardware types providing the basis of cost/performance decisions in the implementation of systems. One could answer questions like how does a K6/233 compare to a P5/233 ? How does a P5/233 compare to a P6/450 ? How much difference is there between an ULTRA ATA disk system and SCSI? Does the difference change with different processor speeds ? Does 1MB cache make a difference with what I want to do ? If implementing a cluster, does saving $2K per box make up for the difference ( providing money for more boxes ) if you use ATA and slightly slower CPUs, rather than higher end platforms ?
I would love to see a single ( and evolving ) location for this kind of info. Hell, I would love to work on compiling it. I have seen some sites with benchmark info, but nothing that seems to try to answer specific questions. The sites that I have seen present specific objective numbers but it was hard to derive any context for the differences between systems.
The problem with overtime is if you get to depend on it. If it goes away, you are hosed! I lived through this a few years ago, loosing about 10 hrs per week of overtime. It was real hard to make ends meet for a while. Of course I seem to be doing the same thing again....savings? yeah, great idea! I will start right after i get that new smp box! oh yeah and a dvd and um hmmm wanna go to Bali next week???
The first cray2 cost about $30 Million. Running a well-optimized code, it churned about 1 Gigaflop. Supercomputers seem to hang at about the same price ( $20-30 M ) but increase performance an order of magnetude or more per generation. This would put is in the range of teraflop machines now - building toward the petaflop for the same $30M. A single 450 MHZ pentium II can do about 70-100 Mflops depending on the exact operation , so ten to fifteen could do the same gigaflop as the cray 2. ( assuming embarrasingly parallel code. ) These computers are easily available for $1K in quantity. 150 should yeld 10GF at under 200k including network, and 1500 should yield 100GF at under $2M. I think this should still look like 10X more cost effective than the massive parallel and vector/mpp machines. If you used dual processor systems, the cost/performance would be even better.
Slashdot Mirror
I simply write all my docs with inverted reality. If my research has shown that the speed of object x is fourteen furlongs per fortnight, I document that the speed is NOT fourteen fpf. Bwaaaaahahahaaa !
Z
What I want to see in a filesystem is a combination of Logical Volume Manager, Journaling and Encryption. This would provide quick booting even with large mounted partitions ( a feature of journaling ) the ability to have a volume larger than a single physical disk drive ( a feature of LVM ) and the security of having the data encrypted. I would think that by integrating these features into a single filesystem there may be some benefit in performance over trying to glue seperate implementations of these features together. There, I did the hard part - creating the idea. Now one of you hacker types, please begin coding !
I wonder if IBM will use RH 7.0 as their release version? That would be great for creating positive first impressions of Linux for corporate types !
In a perfect world, a .0 release of any software could be trusted to be correct. So far, my world is far from perfect. As a sysadmin running production on linux boxen, I would never consider putting a .0 release of ANYONES distribution on a production machine. At the very least, I would wait a few weeks till all the most obvious flaws had been addressed with patches etc. If I were force ^h^h^h^h^h strongly encouraged to use a .0 release somewhere, I would sure as hell put it on a test box as long as possible before moving it onto production. I will not say either good things or bad things about RedHat or any other vendor. As far as I am concerned, like the IRS, they just are. I deal with their existance and try to keep my life easy and stable - I have found ways to use their efforts and products to my advantage and so far have not been burned too bad. When I do have problems, with an honest apprasal of the situation I realize that there was something I could have done ( some might even say should have done ) to avoid the problem in the first place.
I think an important quality of Internet access which is desirable for the evolution of society into cyberspace ( opening myself up a bit there eh? ) is that Internet access must be considered as always on, no cost and fast. It needs to be so cheap and convienient to use that there is simply no thought of not using it for whatever communication / research / play / whatever task is at hand. Ie do not want to continue or exacerbate the class dichotomy of the computer/internet haves vs. the have-nots. By charging per unit access, poor people will have less access to Internet, which will serve to keep them poor. Sounds like bad juju to me.
Datalens seems to be down. Could be something as simple as a telephone guy stealing a pair from their uplink T-1. This can take days to fix, since the telco will say, " no, we havent done anything to YOUR circuit ". I have seen T-1 lines go away for 3 or 4 days from causes like this. How long has the site been down? Does anyone know how big of an ISP Datalens is ?
Will this release permit the development of everyman encryption devices ? One of the common rumors (facts?) of our time is the government intercepting all communication and computer-analysing it for evil words etc. Will someone now make nifty jeez-oh pocket encryptors cheap and easy enough for everyman to use - something to go between a telephone set and the public network, or perhaps between a microphone / speaker and a radio set ? Is the lack of these products in the real world (TM) proof of excessive government controls and insidious dark forces ? And what about Naomi ?
Awww Jeez ! Fixed Link
San Ramon Unified School District in conjunction with Rotary Club sponsored a web programming contest. This contest was for teams of students who would form businesses providing web site design for business customers, who were participating in the contest. The teams were evaluated on both the technical merit of the website and their ability to communicate well with the "client" and extract meaningful knowlege of the business so as to be able to build an effective web site. This Link provides more information on previous contests.
I wonder about the value of overclocking and pushing technology to the extremes in home computers, when the designers of supercomputers have realized that the performance gains achieved by pushing clock speed do not scale linearly with the cost. At some point a maximum speed will be reached and the cost of going any further will be staggering. The architects of high end systems have realized that massively paralell systems were the most practical way to obtain orders of magnatude increases in performance. I think that effort designing languages and programming pardigms to effectively use MPP architecture may be effort better spent than building fluorinert or alcholhol cooled overclocked systems. I suppose these systems are more akin to chopped channeled and supercharged 43 ford roadsters than to a 747 in terms of transporting people and cargo.
Consider an intentionally designed and created virus. And biologically active viruses do go from species to species - most influenza in china comes from birds, particularly chickens.
DNA based computers might be much more sensitive to the effects of radiation, or other mutagenic factors. I can just see the label on the computer, " Caution, the hacker general has determined that smoking causes crashes ".
One interesting tool used in DNA research is called PCR, which is a way of multiplying the amount of DNA in a sample. 24 cycles of PCR amplification would yield a 2 ^ 24 increase in the quantity of a DNA based circuit, and takes less than a day to do. For once, memory could grow faster than data!
I would be concerned about the possibility of your computer infecting you with a biologically active virus....
We have several VA boxes here, and paid extra for "enhanced support" ( one of the joys of working for a very large multi-national corporation - they buy enhanced support for EVERYTHING ). When we had problems using VA supplied gigabit ethernet cards, VA sent two people out here for a couple of days to help us and we did eventually fix the problem.
My perception of VA's linux system support is that they provide a reasonably recent distribution with additional drivers for any uncommon hardware that they provide ( we have a Mylex extreme raid controller ). I think ( and have suggested to them ) that they should provide regular system updates - sending out new CDs with the latest kernels and patches. They could send the cds out quarterly perhaps even with a special update when a new version of whatever distribution ( we got redhat from them ) is released. A newsletter would also be a good idea. They need to go beyond having just good solid products and work on evangalisim, and going the extra mile in customer service is one way to do it. VA systems do cost more than some other companies for the same motherboard / cpu speed / memory configuration. I like their chassis better than the cheaper ones ( Boombox for example has very poor cooling design ). changing from "yes we have customer service " to " customer service is our most important product " would help VA. Many of the things they do - for example employing and supporting the development projects of some of the leaders in linux GUI software, as well as other significant contributors to linux are good things for a company to do. They help the whole linux community, but are probably not recognized by the majority of VA's customers. Having that extra mile customer service would be recognized.
Subject number is customer service for Etoys. I am an E-toys customer ( up untill today ). I called them and the person I spoke with was interested and reasonable and said I would get a call back. Perhaps if more people called, they may realize the scope of discontent with their actions. ( at $.10 per minute LD 800 charges )
Perhaps we should all wear togas ? At any rate, this should be a blast. Rather than the hit or miss chance meetings of fellow posters around a /. booth, this may be a good opportunity to meet others of this here fine community ! !
Z
Hmmm. I wonder what they are going to open beyond what has always been available ? The only things that seem to be hidden are related to domain owner authentication and billing. This may make it easier ( or perhaps only different ) for spam generators to get the contact info for large numbers of domains, but to my knowlege the info has always been there. In fact, I remember getting spammed by someone selling a list of domain managers a while back.
I think the most important aspect of this new chip is the continued pressure of competition on Intel.
This should increase the motivation for the continuation of gains in performance and cost / performance ratio. In terms of what Joe Doakes average hobbiest gets for the money, I think the optimum design right now is a dual processor system using a not-quite bleeding edge CPU - say a PII 450 or therabouts. The cost of adding a second CPU to a system seems to be less than the cost of bumping up the clock another hundred or so MHZ. This does not account for improvements in internal efficiency, but you do get a lot more bang for the buck with 2 CPU's. The more people who have multiprocessor systems, wether SMP, clusters or even shared-memory systems the faster we will see the evolution of new software paradigms that will truly advance the state of the art.
Z
I worked with a client who had 2 IR Wireless links (LACE - Laser Atmospheric Comm. Equipment ? ) installed. The systems were configured for T-1, although they could be used for Ethernet at 10Mbps. The path length was about 1/2 mile in each case. One link was pretty reliable, the other was quite troublesome. In the troublesome link, one end was mounted atop a chimney ( can you say wind? Sure! ) and the other was on a floating membrane roof. The links cost about $37K each, including the T-1 to Fiber converter and the mounting hardware. They were inconvienient to work with due to the mounting locations and hardware, and when aligning them, very minute adjustments of knurled-headed screws made significant changes in signal. The complete range of usefull adjustment was on the order of one rotation of the screw!
I imagine that there are applications where these types of things are usefull, but the installation has to be rock-solid and well engineered.
I think a better growth path is to install fiber. Lots of fiber. Everywhere. NOW !
Z
I, for one would like to see a set of benchmark results for Linux that would help a person to make decisions about the configuration of hardware platforms for systems. I would like to see a set of test results for different types of system activity, such as compiling code, raytracing/graphics/visualization, file system access, network bandwidth, combined network/filesystem access etc. This set of measurements could then be run on a variety of hardware types providing the basis of cost/performance decisions in the implementation of systems. One could answer questions like how does a K6/233 compare to a P5/233 ? How does a P5/233 compare to a P6/450 ? How much difference is there between an ULTRA ATA disk system and SCSI? Does the difference change with different processor speeds ? Does 1MB cache make a difference with what I want to do ?
If implementing a cluster, does saving $2K per box make up for the difference ( providing money for more boxes ) if you use ATA and slightly slower CPUs, rather than higher end platforms ?
I would love to see a single ( and evolving ) location for this kind of info. Hell, I would love to work on compiling it. I have seen some sites with benchmark info, but nothing that seems to try to answer specific questions. The sites that I have seen present specific objective numbers but it was hard to derive any context for the differences between systems.
The problem with overtime is if you get to depend on it. If it goes away, you are hosed! I lived through this a few years ago, loosing about 10 hrs per week of overtime. It was real hard to make ends meet for a while. Of course I seem to be doing the same thing again....savings? yeah, great idea! I will start right after i get that new smp box! oh yeah and a dvd and um hmmm wanna go to Bali next week???
The first cray2 cost about $30 Million. Running a well-optimized code, it churned about 1 Gigaflop. Supercomputers seem to hang at about the same price ( $20-30 M ) but increase performance an order of magnetude or more per generation. This would put is in the range of teraflop machines now - building toward the petaflop for the same $30M. A single 450 MHZ pentium II can do about 70-100 Mflops depending on the exact operation , so ten to fifteen could do the same gigaflop as the cray 2. ( assuming embarrasingly parallel code. ) These computers are easily available for $1K in quantity. 150 should yeld 10GF at under 200k including network, and 1500 should yield 100GF at under $2M. I think this should still look like 10X more cost effective than the massive parallel and vector/mpp
machines. If you used dual processor systems, the cost/performance would be even better.