The msdn.microsoft.com server does not even work unless you are using IE, apparently. I get the following message in Firefox:
An error occurred on the server when processing the URL. Please contact the system administrator.An error occurred on the server when processing the URL. Please contact the system administrator.
I then dropped back to Netscape and tried that and got:
An error occurred on the server when processing the URL. Please contact the system administrator.An error occurred on the server when processing the URL. Please contact the system administrator.
I'm sure this is intentional at Microsoft. And yes, the message came out doubled on each hit, and yes, there was no space between "administrator." and "An error" between them.
Microsoft knows they don't own most of these rights. But what they are trying to trick you (e.g. lots of businesses and their managers and lawyers) into, is effectively agreeing to license everything through them. Once they have a grasp on you with this, they can later force you to sign additional licenses to be able to use upgrades to these protocols, even though you would not need to if you didn't sign this one to begin with. It's a form of entrapment, and the terms are non-negotiable once entrapped. These future licenses can then do evil things varying from costing you more money, to prohibiting you from using software they did not license you to use to make use of the protocols you are licensing from them (e.g. you can't use Linux to talk TCP/IP now because you have agreed to do all your TCP/IP through our licensing).
It's not evil to those who understand what is going on. The vast majority of business people actually have no idea. They will be the gullible ones thinking they are covering their own ass now, when in fact they are handing it over to Microsoft to chew on any time they want to.
If the codecs are separate.so files, then there should be no need to rebuild mplayer. It should simply know where to find the codecs if the original application is designed properly (to the principle of modularity). The whole idea of modules is to be able to add or delete them as needed without rebuilding or reinstalling the whole thing.
Of course, if the codecs are not too large and not too numerous, they could be statically linked into the executable and make it easier on everyone.
I used to work at a major financial services company. This was just as commercialism was just discovering the existance of the internet, so I was hired to design and deploy their high speed redundant connectivity. One thing this company did right, I think, is that all of their security was focused through the VP of Auditing, who reported to the CFO. And the guy who had this position was smart enough to know he knew very little about security and had to learn. I actually got to teach him more about it. We formed a group of people (at my suggestion), including another network engineer, two accountants, and one of the staff lawyers, as the security committee. His original mandate was network security. But in our first group meeting I gave a presentation on one of my long long ago hacking efforts (back in the mainframe days) that successfully broke into a major insurance company's three mainframes. I explained to them how I did it using entirely social engineering. Of course I had knowledge of the system, but I didn't utilize any bugs in the system to get in. With this I was able to get the group to change the focus of security from one strictly focusing on computer technology, to one that would be applied to everything the company did. Software bugs and misconfigured servers are, of course, important, but people are the weakest link in security, and this is even more so the larger a corporation is. Every operation of a company must consider security across the board.
However, this is also an area where a monopoly is least desireable. This is shown by the abuses PayPal has done with their shoddy customer service, and the ripoffs they've done against many people. What it comes down to now is that you either use PayPal or you don't... and a lot of people don't. If a seller is able to take my credit card directly, fine. If not, maybe I'll send them a money order or cashier's check. But I'm not risking my money in PayPal.
I do have an alternative system idea in mind, and I am preparing a document on how it would work. It involves a new exchange system between banks that parallels their existing money exchange systems. An auction seller or shopping site would generate an "open transaction" through their bank (or equivalent service provider). The transaction code is given to the buyer, who then can submit it to their bank for payment. The security is a function of buyer's arrangement with their own bank. Once the bank has determined they are truly dealing with their own account holder who has approved payment on the transaction, payment is sent through a central clearinghouse. It should take under a minute to go through to the seller. Responsibility for fraud would be at the sending bank. The system would also send no private information on the transaction; it would be a blind, and non-reversible, payment. Seller would not know where the payment came from, but would know it cannot be reversed within the system. If the sending bank made a mistake, such as letting an unauthorized person access the buyer's account, that's the responsibility of the sending bank, and buyer's would choose their bank on the basis of how well they deal with security. Both parties would never have to deal with entities other than the one they chose to be their bank.
VNC is a relatively simple protocol that allows integrating it into graphical server programs that have a need to generate their graphics internally, and let users connect to them (something X cannot do because the client and server roles are reversed) to carry on the graphical session. By just using any VNC client, you can connect to such a server directly. The simplicity of VNC allows it to be done without much headache. Does NX work equally simple? Point me to the online specification of the NX protocol and I can see if that is so or not. My first worry is that someone said it uses SSH instead of SSL. That's a major worry right there (VNC, like virtually anything else, can be wrapped in SSL if needed).
When tuning along the dial, these signals can be heard at specific intervals. The interval spacing is the fundamental frequency. Each point is a harmonic. In this case it does not "just happen" to be at 121.5 MHz... it is at 121.5 for a reason, and that is because the 7722nd harmonic of the horizontal sweep frequency is 121.5 MHz.
Which harmonics are stronger does depend on the waveform of the involved signal. A sawtooth is going to have a fast rise and slow decay. And that fast rise time can favor those harmonics that happen to have intervals around where harmonics of a waveform which had both fast rise and fast decay with the same time interval would show up (a higher frequency and this a larger spacing).
The frequency of the NTSC color subcarrier (the TV color system used in analog video standards in North America and Japan) is defined as exactly 5 MHz times 63/88. That works out to 3.579545454.... (infinitely repeating 54's) MHz. The horizontal scanning frequency is then defined as a 2/455 times the color subcarrier frequency. That works out to 15734.26573426.... (infinitely repeating 573426's) Hz (very nearly the original monochrome horizontal frequency of 15750 Hz). This is where the problem lies. 121.5 MHz divided by 7722 is exactly the same frequency as the horizontal in an NTSC color video signal.
The 7722nd harmonic shouldn't really be that strong, right? Usually not. But the harmonics can get to be very strong overall even at such high orders when dealing with modulating the high voltages needed for the horizontal sweep. There should be some low pass filters that prevent that from getting into the VHF range. But if the filters are absent, or were incorrectly installed, or were damaged somehow, and if some wires formed some resonance near 121.5 MHz (like wires going out to cable, speakers, etc)... a wavelength of about 2.47 meters or 8.1 feet... it is possible that harmonic, and a bunch of others near it, could be enhanced and radiated.
The local oscillator in the tuner is a remote possibility. But it would have to be tuned to be receiving a video carrier at 75.75 MHz based on the common satndard of 45.75 MHz for the IF stage in the tuner. But there is no TV broadcast on that frequency in the US... though I could not rule out there being something on that frequency from a cable system. Still, it wouldn't be an expected place for a TV to tune to. But if the TV has a non-standard IF frequency, the local oscillator getting on 121.5 MHz by some expected channel could be possible. Those leak a lot and it's how the snoops can tell what channel you are tuned to by spying on the RF emitted from your house.
If just this one TV had the problem, then apparently it must be a manufacturing defect or shipping damage (or maybe user damage or tampering). If it were a design problem, I'd sure we'd hear more about it. That probably rules out the CPU clock frequency.
This is true. But PayPal may be making arrangements with its bank(s) to process the checks between them electronically, to speed things up for both. That's just speculation since I have no idea what is really going on, but it fits what we see happening just a bit better.
It seems PayPal is trying to do as many things as it can to be like a bank, without falling under banking law jurisdiction. But in their zeal, maybe they will step too far and we can through some court force them to be a real bank. Then it might become safe to use. If a real bank were to start doing some innovative internet services, I think lots of things would move forward. But I think too many bank executives are still struggling to figure out what the extra button on the mouse is for.
This is the "Check Clearing for the 21st Century Act", commonly known as "Check 21". Basically it allows a bank anywhere along the path from where the check is first deposited, to its arrival for payment at your bank, to replace the paper check with a front and back image scan. The law provides that your copy of this substituted check must be treated like an original check for the purpose of things like using it as a receipt to show you paid. For example, if your landlord failed to record the fact that you paid the rent, but deposited your check anyway, the law requires this substitute check image (printed back to you by your bank) be accepted as proof the check was deposited just as the original would be.
Banks are not required to do the image scan of checks, but they are allowed to do so. Banks are required to accept the image scan in place of those checks when the image scan gets done. If PayPal is allowing you to write checks against your account (but they would BE a bank if this happens, I'd think), they would have to update their software by October 28 to comply. More likely, if "Check 21" is an issue here, is that they may be adding some software to allow them to image scan checks made as payment to them. But the more they do like this, the closer they become to being a real bank.
When an image scan is done, the check can be processed much faster because it can now be sent to the account holder bank electronically. This is where the "float" many people depend on can start to disappear. OTOH, your bank may be able to get funds into your account for checks paid to you that you deposit equally faster. There is a possibility that PayPal was doing things that depend on the "float". Many business and people have been doing that for years. Practices will now have to change.
RAM occupancy that can cause other processes to have more pages swapped out than otherwise would be, slowing them down to swap them back in when needed
More segement table switching and page translation cache flushing than would otherwise be, when process VMs are switched
More CPU gate state changes happening, especially in numeric and floating point circuitry, using more electrical power and dissipating more heat
Periodic use of additional bandwidth to return results and fetch a new work unit
If you worked for me, and I discovered you were running SETI on one or more servers, you'd be in for this lecture, and a reprimand instructing you to never do that again on any computer you work on under my authority. I wouldn't fire on a first offense, but a second offense opens that possibility.
However, this kind of violation normally should not rise to the level of a firing (not considering the slander his boss committed). It may have simply been the last in a series of violations; we don't really know (and his boss would be out of place spelling those out in the media, as well). So maybe in the broader picture, the time for a firing had come.
What I think should be done is reprimand (with a financial penalty) the boss in this case for the slander that is done. Then there should be an investigation to verify if the firing was justified (it may have been).
Mr. Hayes certainly sounds like someone who has a chip on his shoulders, based on the way he made these remarks. I'm sure this incident will result in a lot of resumes being sent in by others willing to take the job (and leave SETI@home at home). But mine won't be among them (even though it's the kind of job, and location, I could do) because I do not like working for unprofessional people.
I've actually gotten calls from headhunters again. The jobs are crap, but at least they are starting to call. But these calls have also given me a clue about what the cause is, and this can influence the (possibly temporary) rise in pay levels.
That cause is the current H-1B cap.
On October 1, 2003, the limit on the number of new H-1B visas that can be issued each year fell back to 65,000 after Congress declined to renew the 3 year raise on that limit. By January 2004, the fiscal year quota had been met. That suggests that American businesses would have, if they could have, hired as many as 260,000 foreign nationals to fill jobs they don't want to hire Americans for. But with the low cap, they are now forced to do just that: hire Americans.
The coming danger is that the Bush Administration, and the dominant party in Congress, want to raise the cap again, presumably so they and their other fat cat friends can enjoy the riches of investing in businesses that grow by selling out American to cheap foreign labor. When this happens, watch for the gains made by IT workers to evaporate.
Free Trade? That's what they like to call it. Free maybe. Fair and level, certainly not. When the monetary exchange rate is so distorted as it is, obviously intended to favor rich Americans, someone making what would be the poverty line in the US can live quite well in India. They wouldn't have a big palace or anything like that, but they could have a nice (well air conditioned... that's needed in India) 2 to 4 bedroom modern apartment in a clean neighborhood in a major Indian city like Mumbai or Bangalore. So of course they can work for less, but it isn't less to them, it's more.
The blame lies not with Indians, though (they are merely doing better for themselves as anyone would want to do), but with the world bankers and the American politicians (Republicans as well as Democrats who have some of the most clueless politicians around) intent on screwing the lower 99% to make themselves and their friend richer.
By the way, one of the jobs I got a call about was a contract gig for Accenture, a formerly American company that decided it didn't want to pay taxes anymore and moved their company to the Bahamas. But the US government still grants them many lucrative contracts, which they would like to fulfill with cheaper non-American labor as soon as Bush gets that cap raised.
One aspect of fighting spam is fighting the costs that spamming imposes. This is one of the reasons I don't generally use content based analysis. But, it would be OK to also do these checks at the RFCx822 level for mail that wasn't rejected before, that would otherwise be accepted anyway.
It's so easy to add an SPF record because it uses an already existing DNS standard TXT record, that if any DNS hoster can't support, it's time to change to another one.
Slashdot Mirror
The msdn.microsoft.com server does not even work unless you are using IE, apparently. I get the following message in Firefox:
I then dropped back to Netscape and tried that and got:
I'm sure this is intentional at Microsoft. And yes, the message came out doubled on each hit, and yes, there was no space between "administrator." and "An error" between them.
Oh wait ... Microsoft has been Slashdotted!
Microsoft knows they don't own most of these rights. But what they are trying to trick you (e.g. lots of businesses and their managers and lawyers) into, is effectively agreeing to license everything through them. Once they have a grasp on you with this, they can later force you to sign additional licenses to be able to use upgrades to these protocols, even though you would not need to if you didn't sign this one to begin with. It's a form of entrapment, and the terms are non-negotiable once entrapped. These future licenses can then do evil things varying from costing you more money, to prohibiting you from using software they did not license you to use to make use of the protocols you are licensing from them (e.g. you can't use Linux to talk TCP/IP now because you have agreed to do all your TCP/IP through our licensing).
It's not evil to those who understand what is going on. The vast majority of business people actually have no idea. They will be the gullible ones thinking they are covering their own ass now, when in fact they are handing it over to Microsoft to chew on any time they want to.
So how is the PPC port coming along? I was hoping it would make it into 5.3.
If the codecs are separate .so files, then there should be no need to rebuild mplayer. It should simply know where to find the codecs if the original application is designed properly (to the principle of modularity). The whole idea of modules is to be able to add or delete them as needed without rebuilding or reinstalling the whole thing.
Of course, if the codecs are not too large and not too numerous, they could be statically linked into the executable and make it easier on everyone.
I used to work at a major financial services company. This was just as commercialism was just discovering the existance of the internet, so I was hired to design and deploy their high speed redundant connectivity. One thing this company did right, I think, is that all of their security was focused through the VP of Auditing, who reported to the CFO. And the guy who had this position was smart enough to know he knew very little about security and had to learn. I actually got to teach him more about it. We formed a group of people (at my suggestion), including another network engineer, two accountants, and one of the staff lawyers, as the security committee. His original mandate was network security. But in our first group meeting I gave a presentation on one of my long long ago hacking efforts (back in the mainframe days) that successfully broke into a major insurance company's three mainframes. I explained to them how I did it using entirely social engineering. Of course I had knowledge of the system, but I didn't utilize any bugs in the system to get in. With this I was able to get the group to change the focus of security from one strictly focusing on computer technology, to one that would be applied to everything the company did. Software bugs and misconfigured servers are, of course, important, but people are the weakest link in security, and this is even more so the larger a corporation is. Every operation of a company must consider security across the board.
SLS (Softlanding Linux System) was the first. And it led to both Slackware and Debian.
Send this reply back to them:
This is an automatically generated Delivery Status Notification.
Delivery to the following recipient(s) failed.
Reason code:
For additional assistance, contact your ISP or email administrator.
So why not just use R-718 refrigerant? It works just like water.
Actually not. In the deeper details there is a way to monitor for that. It's just the receipient of the money who won't get private information.
However, this is also an area where a monopoly is least desireable. This is shown by the abuses PayPal has done with their shoddy customer service, and the ripoffs they've done against many people. What it comes down to now is that you either use PayPal or you don't ... and a lot of people don't. If a seller is able to take my credit card directly, fine. If not, maybe I'll send them a money order or cashier's check. But I'm not risking my money in PayPal.
I do have an alternative system idea in mind, and I am preparing a document on how it would work. It involves a new exchange system between banks that parallels their existing money exchange systems. An auction seller or shopping site would generate an "open transaction" through their bank (or equivalent service provider). The transaction code is given to the buyer, who then can submit it to their bank for payment. The security is a function of buyer's arrangement with their own bank. Once the bank has determined they are truly dealing with their own account holder who has approved payment on the transaction, payment is sent through a central clearinghouse. It should take under a minute to go through to the seller. Responsibility for fraud would be at the sending bank. The system would also send no private information on the transaction; it would be a blind, and non-reversible, payment. Seller would not know where the payment came from, but would know it cannot be reversed within the system. If the sending bank made a mistake, such as letting an unauthorized person access the buyer's account, that's the responsibility of the sending bank, and buyer's would choose their bank on the basis of how well they deal with security. Both parties would never have to deal with entities other than the one they chose to be their bank.
VNC is a relatively simple protocol that allows integrating it into graphical server programs that have a need to generate their graphics internally, and let users connect to them (something X cannot do because the client and server roles are reversed) to carry on the graphical session. By just using any VNC client, you can connect to such a server directly. The simplicity of VNC allows it to be done without much headache. Does NX work equally simple? Point me to the online specification of the NX protocol and I can see if that is so or not. My first worry is that someone said it uses SSH instead of SSL. That's a major worry right there (VNC, like virtually anything else, can be wrapped in SSL if needed).
When tuning along the dial, these signals can be heard at specific intervals. The interval spacing is the fundamental frequency. Each point is a harmonic. In this case it does not "just happen" to be at 121.5 MHz ... it is at 121.5 for a reason, and that is because the 7722nd harmonic of the horizontal sweep frequency is 121.5 MHz.
Which harmonics are stronger does depend on the waveform of the involved signal. A sawtooth is going to have a fast rise and slow decay. And that fast rise time can favor those harmonics that happen to have intervals around where harmonics of a waveform which had both fast rise and fast decay with the same time interval would show up (a higher frequency and this a larger spacing).
The frequency of the NTSC color subcarrier (the TV color system used in analog video standards in North America and Japan) is defined as exactly 5 MHz times 63/88. That works out to 3.579545454.... (infinitely repeating 54's) MHz. The horizontal scanning frequency is then defined as a 2/455 times the color subcarrier frequency. That works out to 15734.26573426.... (infinitely repeating 573426's) Hz (very nearly the original monochrome horizontal frequency of 15750 Hz). This is where the problem lies. 121.5 MHz divided by 7722 is exactly the same frequency as the horizontal in an NTSC color video signal.
The 7722nd harmonic shouldn't really be that strong, right? Usually not. But the harmonics can get to be very strong overall even at such high orders when dealing with modulating the high voltages needed for the horizontal sweep. There should be some low pass filters that prevent that from getting into the VHF range. But if the filters are absent, or were incorrectly installed, or were damaged somehow, and if some wires formed some resonance near 121.5 MHz (like wires going out to cable, speakers, etc) ... a wavelength of about 2.47 meters or 8.1 feet ... it is possible that harmonic, and a bunch of others near it, could be enhanced and radiated.
The local oscillator in the tuner is a remote possibility. But it would have to be tuned to be receiving a video carrier at 75.75 MHz based on the common satndard of 45.75 MHz for the IF stage in the tuner. But there is no TV broadcast on that frequency in the US ... though I could not rule out there being something on that frequency from a cable system. Still, it wouldn't be an expected place for a TV to tune to. But if the TV has a non-standard IF frequency, the local oscillator getting on 121.5 MHz by some expected channel could be possible. Those leak a lot and it's how the snoops can tell what channel you are tuned to by spying on the RF emitted from your house.
If just this one TV had the problem, then apparently it must be a manufacturing defect or shipping damage (or maybe user damage or tampering). If it were a design problem, I'd sure we'd hear more about it. That probably rules out the CPU clock frequency.
This is true. But PayPal may be making arrangements with its bank(s) to process the checks between them electronically, to speed things up for both. That's just speculation since I have no idea what is really going on, but it fits what we see happening just a bit better.
It seems PayPal is trying to do as many things as it can to be like a bank, without falling under banking law jurisdiction. But in their zeal, maybe they will step too far and we can through some court force them to be a real bank. Then it might become safe to use. If a real bank were to start doing some innovative internet services, I think lots of things would move forward. But I think too many bank executives are still struggling to figure out what the extra button on the mouse is for.
This is the "Check Clearing for the 21st Century Act", commonly known as "Check 21". Basically it allows a bank anywhere along the path from where the check is first deposited, to its arrival for payment at your bank, to replace the paper check with a front and back image scan. The law provides that your copy of this substituted check must be treated like an original check for the purpose of things like using it as a receipt to show you paid. For example, if your landlord failed to record the fact that you paid the rent, but deposited your check anyway, the law requires this substitute check image (printed back to you by your bank) be accepted as proof the check was deposited just as the original would be.
Banks are not required to do the image scan of checks, but they are allowed to do so. Banks are required to accept the image scan in place of those checks when the image scan gets done. If PayPal is allowing you to write checks against your account (but they would BE a bank if this happens, I'd think), they would have to update their software by October 28 to comply. More likely, if "Check 21" is an issue here, is that they may be adding some software to allow them to image scan checks made as payment to them. But the more they do like this, the closer they become to being a real bank.
When an image scan is done, the check can be processed much faster because it can now be sent to the account holder bank electronically. This is where the "float" many people depend on can start to disappear. OTOH, your bank may be able to get funds into your account for checks paid to you that you deposit equally faster. There is a possibility that PayPal was doing things that depend on the "float". Many business and people have been doing that for years. Practices will now have to change.
For more information:
If you worked for me, and I discovered you were running SETI on one or more servers, you'd be in for this lecture, and a reprimand instructing you to never do that again on any computer you work on under my authority. I wouldn't fire on a first offense, but a second offense opens that possibility.
However, this kind of violation normally should not rise to the level of a firing (not considering the slander his boss committed). It may have simply been the last in a series of violations; we don't really know (and his boss would be out of place spelling those out in the media, as well). So maybe in the broader picture, the time for a firing had come.
What I think should be done is reprimand (with a financial penalty) the boss in this case for the slander that is done. Then there should be an investigation to verify if the firing was justified (it may have been).
Mr. Hayes certainly sounds like someone who has a chip on his shoulders, based on the way he made these remarks. I'm sure this incident will result in a lot of resumes being sent in by others willing to take the job (and leave SETI@home at home). But mine won't be among them (even though it's the kind of job, and location, I could do) because I do not like working for unprofessional people.
There are three options in the design of a system:
- How the boss would do it
- How the developer would do it
- The way that actually works
Choose two.I've actually gotten calls from headhunters again. The jobs are crap, but at least they are starting to call. But these calls have also given me a clue about what the cause is, and this can influence the (possibly temporary) rise in pay levels.
That cause is the current H-1B cap.
On October 1, 2003, the limit on the number of new H-1B visas that can be issued each year fell back to 65,000 after Congress declined to renew the 3 year raise on that limit. By January 2004, the fiscal year quota had been met. That suggests that American businesses would have, if they could have, hired as many as 260,000 foreign nationals to fill jobs they don't want to hire Americans for. But with the low cap, they are now forced to do just that: hire Americans.
The coming danger is that the Bush Administration, and the dominant party in Congress, want to raise the cap again, presumably so they and their other fat cat friends can enjoy the riches of investing in businesses that grow by selling out American to cheap foreign labor. When this happens, watch for the gains made by IT workers to evaporate.
Free Trade? That's what they like to call it. Free maybe. Fair and level, certainly not. When the monetary exchange rate is so distorted as it is, obviously intended to favor rich Americans, someone making what would be the poverty line in the US can live quite well in India. They wouldn't have a big palace or anything like that, but they could have a nice (well air conditioned ... that's needed in India) 2 to 4 bedroom modern apartment in a clean neighborhood in a major Indian city like Mumbai or Bangalore. So of course they can work for less, but it isn't less to them, it's more.
The blame lies not with Indians, though (they are merely doing better for themselves as anyone would want to do), but with the world bankers and the American politicians (Republicans as well as Democrats who have some of the most clueless politicians around) intent on screwing the lower 99% to make themselves and their friend richer.
By the way, one of the jobs I got a call about was a contract gig for Accenture, a formerly American company that decided it didn't want to pay taxes anymore and moved their company to the Bahamas. But the US government still grants them many lucrative contracts, which they would like to fulfill with cheaper non-American labor as soon as Bush gets that cap raised.
Does the source code count if I download it and build my own binary?
I have something better than ROT-13. It's Triple-ROT-13!
I hope they do. That would mean less cash available to spend on lawyers.
One aspect of fighting spam is fighting the costs that spamming imposes. This is one of the reasons I don't generally use content based analysis. But, it would be OK to also do these checks at the RFCx822 level for mail that wasn't rejected before, that would otherwise be accepted anyway.
It's so easy to add an SPF record because it uses an already existing DNS standard TXT record, that if any DNS hoster can't support, it's time to change to another one.