I first used an email like system in 1980 on an IBM mainframe. I was referred to as mail, but not "email". I think at best he might be able to say he was the first to coin the silly term "email". I see no more reason to use the "e" as we don't refer to the network as "electronic". No doubt others might already be using words like "iMail" (Steve, is that you?) or "cmail" or "nmail".
Who cares, though. It all became worthless as soon as spam (all lower case) was invented a few days afterwards.
It's good to know that at least SOME legislators somewhere believe there is an issue with (what I call) a government capitulating to the desires of the terrorists to ruin our society.
Once the bank of cells does go below the detection level, it is still possible for them to still discharge further, depending on the condition. But they should wire the battery pack management system in such a way that it can be powered itself by other means than just the battery pack it manages. When the charge connection is plugged in, it should be made so that a low level of power is provided to bring the management system up without it being powered from the pack. At this point, it should be able to test the battery pack condition to determine if it can be charged back up safely or not. If Tesla didn't think of it, their bad.
Maybe little brats like you should read what was written. I said nothing whatsoever about expecting these dumb people to harden their computers. Of course it is almost impossible. I do in fact know that. And that is exactly why I stated what I did. And that is the lesson these dumb people need to learn. They need to learn that their choice of getting a computer that the vendor has not already hardened is what is causing problems not only for them, but also for everyone else. If the buyer side of the market wanted a hardened computer, the seller side of the market would provide one. We need to be putting intense pressure on the buyer side of the market to do no less. Then vendors like Microsoft would find their market share seriously declining unless they respond to the market demands that need to be happening.
.
It is NOT my role to go after Microsoft. It is not Microsoft that has directly caused me harm. It is the people that choose (way too often just by default) to use weakly coded crapware Microsoft makes that are the problem. These people need to be able to choose between taking the issue up with Microsoft and to their chairman or their chair man and demanding things be fixed... or got to get another OS.
Serving valid DNS data to allow access to sites like virus checkers/removers, and the OS providers (I have a very good idea which one that is), makes it easier for home and small business users to get their computers cleaned up. However, they SHOULD make OTHER sites just go to a page that tells them their computer is infected with a virus that interferes with the computer's ability to locate web sites on the internet. It will be a LONG time getting them all cleaned up otherwise.
... is track down the owners of these computers and charge them ALL with the misdemeanor aiding and abetting cybercrimes. Let's put the blame where it belongs... on dumb people who allow their computers to be infected. In this case, since there was no damage by these owners to others, it can be a misdemeanor. But if it did involve damage to others, then it should be a felony charge.
If the chip doesn't work at all, how is scanning it going to get a "valid" status from the government database? If your interpretation prevails, then the system fails because all anyone would need to do to game the system is to kill the chip. Instead, we need more reliable chips and a better access to a fallback verification system (for which the chip should be merely a means to expedite validation).
And how would you do it w/o one thread or process per client? Multiplexing?
We actually need multiple web server models. Processes are needed for multi-user environments for security reasons. But that impacts performance for single-user environments. OTOH, Apache does process based stuff all wrong, too (they do execve() for CGI and that is where the performance dies).
Anonymous Coward switching to NginX would prove software can improve. The question is, does it need to be done by rethinking and starting over from scratch.
The issue is why it is that loading programs take much longer than it takes for a simple program to just read as much data from a file that is essential for the big program to get to a point where it can present the user experience. I think what the author of the blog really means is that the program should just do the minimum needed to get the real starting screen up first, make user input functional, and then load the rest in the background, starting with the subscreens program sections the user typically uses first.
What actually is happening between when the splash screen comes up, and when the user can start to use the application? Since there is no user input, yet, it would have to be basically the same thing happening each time it starts. So why not just save the memory image of all that the first time, and just map it in every time after that? Of course that means thinking about your programming environment in a different way, and very likely doing some major modifications of that, so that things like building all kinds of data structures to hold all kinds of discrete info like icon images, can be just mapped in from prior builds rather than reading hundreds of individual files each time the program starts. See, the problem here is that people are still thinking in terms of starting up a program in a nearly empty virtual memory space. Instead, have all this information preconstructed in a blob. If the blob is missing, or marked invalid, then do rebuild it. But if it is OK, then just map it into memory and move on. This blob will, of course, need to be built in a way that allows it to work no matter what virtual memory address it is loaded at. Unfortunately, a lot of programmers these days don't even know what a memory address is, and think virtual memory is some means to eliminate needing to know that.
They were not fighting the Germans. They were fighting the Nazis. It just happened to be Germans who were the largest group under Nazi control at the time. The Nazis, had they been allowed to stay in power, would be a LOT WORSE than what the UK, USA, and some other countries are currently doing to rip off people's rights. Not that it is impossible for UK and USA to get as bad as the Nazis as this certainly could happen, and is the current direction of movement. But this proposal, while a big step in the direction of Nazism, is not there, yet. This kind of thing needs to be stopped as far from being Nazism as possible. Just replace "Germans" with "Nazis" and that will fix what you posted.
Employers cannot tell me what to do in my private life with my own name (besides telling me not to do illegal things or things that do involve the company). My job function would never be that kind of thing. Maybe some kinds of jobs would need it (TV news personality, HR investigator, etc... but I don't do those kinds of jobs). I did set up a Skype account for teleconferencing, but that was in the company name, not my own (the company owns it, not me). But my company has no rights to my identity. Firing me over this means I have to pick which one of my 5 lawyers friends is going to make a boatload.
Any HONEST and HONORABLE institution should have nothing to hide. With all their working documents publicly available, this kind of thing would not happen. I have to conclude that Heartland does not meet this ideal.
Yes and no. The over-complexity of the x86 architecture is a factor that makes hitting this limit easier. If they had put all the effort into ARM or another RISC architecture that they did put into x86, the end result would be a leaner faster CPU. For 64-bit, we do have PPC64 as an architecture choice. Or if Intel had switched over to ARM long ago, we might already have ARM64 by now. Anyway, a RISC architecture like ARM could go further than x86. I don't know if it could make it to 15 GHz. But I do believe it could get to 8 GHz on the much smaller dies needed for ONE core of ARM vs ONE core of x86. There is digital circuitry doing above 4 GHz for years.
I have not looked at how the CPUs are being wired and made for years. But what I did see back when I did look (around the time the i386 was rumored) worried me. In those days, the circuit lines were unbalanced on the chip. That is, signal flow was strictly monopolar over grounding. Inductance would force current to flow in the substrate to equalize the fields. What is needed is balanced bipolar signal flow, much like you have with ethernet (but the twisting is not essential) to reduce those effects (by confining the EMF to a smaller space and reducing inductive effects). That and, of course, time to propagate (so the chips do need to be made smaller).
I suspect the real technology that will give CPUs an eventual boost is nanowaveguides in vacuum. But they'll have to figure it out.
The codes are printed where the printer knows no yellow is. But printing a solid yellow background will help... by running the printer out of yellow ink.
Kill the yellow channel on the color inks. If your printer uses separate black ink, just leave the color out. If all else fails, print a lot of pages of all-yellow on the same set of several sheets until the yellow ink runs out.
Slashdot Mirror
I first used an email like system in 1980 on an IBM mainframe. I was referred to as mail, but not "email". I think at best he might be able to say he was the first to coin the silly term "email". I see no more reason to use the "e" as we don't refer to the network as "electronic". No doubt others might already be using words like "iMail" (Steve, is that you?) or "cmail" or "nmail".
Who cares, though. It all became worthless as soon as spam (all lower case) was invented a few days afterwards.
It's good to know that at least SOME legislators somewhere believe there is an issue with (what I call) a government capitulating to the desires of the terrorists to ruin our society.
Once the bank of cells does go below the detection level, it is still possible for them to still discharge further, depending on the condition. But they should wire the battery pack management system in such a way that it can be powered itself by other means than just the battery pack it manages. When the charge connection is plugged in, it should be made so that a low level of power is provided to bring the management system up without it being powered from the pack. At this point, it should be able to test the battery pack condition to determine if it can be charged back up safely or not. If Tesla didn't think of it, their bad.
Maybe little brats like you should read what was written. I said nothing whatsoever about expecting these dumb people to harden their computers. Of course it is almost impossible. I do in fact know that. And that is exactly why I stated what I did. And that is the lesson these dumb people need to learn. They need to learn that their choice of getting a computer that the vendor has not already hardened is what is causing problems not only for them, but also for everyone else. If the buyer side of the market wanted a hardened computer, the seller side of the market would provide one. We need to be putting intense pressure on the buyer side of the market to do no less. Then vendors like Microsoft would find their market share seriously declining unless they respond to the market demands that need to be happening.
.
It is NOT my role to go after Microsoft. It is not Microsoft that has directly caused me harm. It is the people that choose (way too often just by default) to use weakly coded crapware Microsoft makes that are the problem. These people need to be able to choose between taking the issue up with Microsoft and to their chairman or their chair man and demanding things be fixed ... or got to get another OS.
Serving valid DNS data to allow access to sites like virus checkers/removers, and the OS providers (I have a very good idea which one that is), makes it easier for home and small business users to get their computers cleaned up. However, they SHOULD make OTHER sites just go to a page that tells them their computer is infected with a virus that interferes with the computer's ability to locate web sites on the internet. It will be a LONG time getting them all cleaned up otherwise.
... is track down the owners of these computers and charge them ALL with the misdemeanor aiding and abetting cybercrimes. Let's put the blame where it belongs ... on dumb people who allow their computers to be infected. In this case, since there was no damage by these owners to others, it can be a misdemeanor. But if it did involve damage to others, then it should be a felony charge.
If the chip doesn't work at all, how is scanning it going to get a "valid" status from the government database? If your interpretation prevails, then the system fails because all anyone would need to do to game the system is to kill the chip. Instead, we need more reliable chips and a better access to a fallback verification system (for which the chip should be merely a means to expedite validation).
And how would you do it w/o one thread or process per client? Multiplexing?
We actually need multiple web server models. Processes are needed for multi-user environments for security reasons. But that impacts performance for single-user environments. OTOH, Apache does process based stuff all wrong, too (they do execve() for CGI and that is where the performance dies).
Anonymous Coward switching to NginX would prove software can improve. The question is, does it need to be done by rethinking and starting over from scratch.
I'd rather have better control features, such as completely redoing the vhost matching method.
The issue is why it is that loading programs take much longer than it takes for a simple program to just read as much data from a file that is essential for the big program to get to a point where it can present the user experience. I think what the author of the blog really means is that the program should just do the minimum needed to get the real starting screen up first, make user input functional, and then load the rest in the background, starting with the subscreens program sections the user typically uses first.
What actually is happening between when the splash screen comes up, and when the user can start to use the application? Since there is no user input, yet, it would have to be basically the same thing happening each time it starts. So why not just save the memory image of all that the first time, and just map it in every time after that? Of course that means thinking about your programming environment in a different way, and very likely doing some major modifications of that, so that things like building all kinds of data structures to hold all kinds of discrete info like icon images, can be just mapped in from prior builds rather than reading hundreds of individual files each time the program starts. See, the problem here is that people are still thinking in terms of starting up a program in a nearly empty virtual memory space. Instead, have all this information preconstructed in a blob. If the blob is missing, or marked invalid, then do rebuild it. But if it is OK, then just map it into memory and move on. This blob will, of course, need to be built in a way that allows it to work no matter what virtual memory address it is loaded at. Unfortunately, a lot of programmers these days don't even know what a memory address is, and think virtual memory is some means to eliminate needing to know that.
They were not fighting the Germans. They were fighting the Nazis. It just happened to be Germans who were the largest group under Nazi control at the time. The Nazis, had they been allowed to stay in power, would be a LOT WORSE than what the UK, USA, and some other countries are currently doing to rip off people's rights. Not that it is impossible for UK and USA to get as bad as the Nazis as this certainly could happen, and is the current direction of movement. But this proposal, while a big step in the direction of Nazism, is not there, yet. This kind of thing needs to be stopped as far from being Nazism as possible. Just replace "Germans" with "Nazis" and that will fix what you posted.
Employers cannot tell me what to do in my private life with my own name (besides telling me not to do illegal things or things that do involve the company). My job function would never be that kind of thing. Maybe some kinds of jobs would need it (TV news personality, HR investigator, etc ... but I don't do those kinds of jobs). I did set up a Skype account for teleconferencing, but that was in the company name, not my own (the company owns it, not me). But my company has no rights to my identity. Firing me over this means I have to pick which one of my 5 lawyers friends is going to make a boatload.
I always wondered where Anonymous Coward works.
Any HONEST and HONORABLE institution should have nothing to hide. With all their working documents publicly available, this kind of thing would not happen. I have to conclude that Heartland does not meet this ideal.
Use of encryption, and international email services on HTTPS, has started to rise in Canada.
Just tweet the links, instead.
Yes and no. The over-complexity of the x86 architecture is a factor that makes hitting this limit easier. If they had put all the effort into ARM or another RISC architecture that they did put into x86, the end result would be a leaner faster CPU. For 64-bit, we do have PPC64 as an architecture choice. Or if Intel had switched over to ARM long ago, we might already have ARM64 by now. Anyway, a RISC architecture like ARM could go further than x86. I don't know if it could make it to 15 GHz. But I do believe it could get to 8 GHz on the much smaller dies needed for ONE core of ARM vs ONE core of x86. There is digital circuitry doing above 4 GHz for years.
I have not looked at how the CPUs are being wired and made for years. But what I did see back when I did look (around the time the i386 was rumored) worried me. In those days, the circuit lines were unbalanced on the chip. That is, signal flow was strictly monopolar over grounding. Inductance would force current to flow in the substrate to equalize the fields. What is needed is balanced bipolar signal flow, much like you have with ethernet (but the twisting is not essential) to reduce those effects (by confining the EMF to a smaller space and reducing inductive effects). That and, of course, time to propagate (so the chips do need to be made smaller).
I suspect the real technology that will give CPUs an eventual boost is nanowaveguides in vacuum. But they'll have to figure it out.
You didn't try any Supermicro boards?
Buy one that has run out of yellow ink.
The codes are printed where the printer knows no yellow is. But printing a solid yellow background will help ... by running the printer out of yellow ink.
Kill the yellow channel on the color inks. If your printer uses separate black ink, just leave the color out. If all else fails, print a lot of pages of all-yellow on the same set of several sheets until the yellow ink runs out.
Or other Scandinavian countries.
Damn. I wish I had mods points today. I blew them all yesterday. Even if I did, I would only be able to do +1. Shucks.