It is inappropriate to ask "how likely was it for this server to have crashed", when there is nothing special about this server other than it is the one we already know has crashed. Once we carefully pick this crashed server from the set of all servers in the world, the "null hypothesis" is gone.
The correct question is "how likely was it for any server to have crashed", and I think the answer is that it was fairly likely.
It is unclear if being New Year's made a crash more likely, but I think we can agree it made it far more likely that we would hear about it.
Assuming your math is correct (I didn't bother to check it), that would be the odds of a random server failing in a random 20 second interval.
You didn't pick a random server, you picked one already known to have crashed. And you didn't pick a random 20 second interval either. The odds of that server crashing in that 20 second interval was 100%, because it was already known to have happened.
This is a classic mis-application of statistics.
Admittedly, the interval right at New Year's is a bit suspicious, since there is some specific code to handle leap years, etc. But given that there wasn't a rash of outages reported, I am going along with the coincidence theory.
Seriously, what kind of ASSHOLE CHAUVINISTIC PIG would say that the person alleged of sexual assault should it not be true would be WORSE off than the victim were it true?
Read it again. He didn't say that at all. He said that the embarrassment of being publicly identified as a victim of sexual assault is less than the embarrassment of being publicly accused of being the assaulter. He was comparing the results of publication, not of the crime itself.
Hrm-- do you have a reference for the discrete math book? I have two (just completed a discrete math course), and IIRC neither talk much about encryption beyond simple XOR ciphers.
Actually, OTP does use a simple XOR cipher. Given a truly random key the same length as the message, that is perfectly adequate.
But going out and trying to sell the information that is inside the books is taking it too far.
No it's not, factual information isn't and shouldn't be protected.
By that logic, a verbatim copy of the whole series would be "factual information" about the contents of the series. This is not like trying to copyright the population of the Congo, or trigonometric tables. This is an original work of fiction, and has no independent existence outside of her work. Based on the description of the lexicon, calling it fair use seems something of a stretch, and calling it "factual information" is a complete farce.
IANAL, but a bit of googling found references to section 110(5) of the Music Licensing Act of 1998. The best summary of the legalese I found was:
Now the Music Licensing Act draws the line between private and public in terms of the type of public establishment, the size, and the stereo equipment used. Restaurants and bars under 3,750 square feet or retail establishments under 2,000 square feet are exempt from paying fees for playing radio or TV broadcasts for their customers. Public places of any size that play radio or TV broadcasts are exempt from paying fees if they use no more than six external speakers (not more than four speakers in each room) for playing music. Public places that play CDs or hire live musicians (that play cover songs or copy songs) are still subject to being licensed for fees.
You can lose the private home exemption and be subject to a license if you charge anyone admission fees to listen to music.
Also, I have often heard that a favorite tactic of the Muzak salesmen is to completely ignore this, and try to convince shop owners they find playing the radio that they are violating copyright, and try to get them to sign up to avoid liability. Apparently, this constitutes fraud, but they are rarely called on it. I don't know if the BMI/ASCAP people do the same sorts of things or not.
I hope this case gets slapped down hard by the UK court system, this is pretty ridiculous.
Can anyone with a sub give us the exact wording/page number, please?
Here it is, from page 4:
This is the last issue of Sys Admin magazine that you will receive. The magazine is ceasing publication as of this issue.
The is often a large measure of regret and sadness when a long-term relations ends, and I feel these emotions now at the end of my relationship with Sys Admin. No other publication really does what Sys Admin does, but you probably already know that. You might not know, however, the challenges currently facing many print publications, particularly small niche ones like Sys Admin. These challenges, which have contributed to the decision to stop publishing the magazine, include circulation woes, online competition, and market shift. I take some small comfort in the fact that Sys Admin fought these battles and survived much longer than many others in similar circumstances.
I have worked for Sys Admin magazine for almost 12 years, and I've had a great time. So, in this note, I need to mention some of the many people who have made working for Sys Admin such a rewarding experience for me. In no particular order, I thank Hal Pomeranze, Joe Casad, Ralph Barker, and Robert Ward for making me seem smarter than I really am. I thank Rikki Endsley Kite for therapy shopping, making me do things that scare me, and being a fabulous friend. I thank Lori White and Twyla Watson Bogaard for always reminding me that life goes down better with humor. I thank the regular columnists and writers for their loyalty and consistently excellent articles, and I thank all the readers and contributors for making Sys Admin as successful as it was for as long as it was.
Early on, Apple put a lot of research into their human interface guidelines. They found that, in general, the amount of time it takes a user to click a button, menu, or other control was inversely proportional to the size of the control. The exception was items at the edge of the screen, which were effectively "infinitely large" in one direction (since it was impossible to overshoot). Thus, in many cases, putting menus at a screen edge can make them faster to access. I find that, if my trackpad sensitivity is set properly, I can hit menus at the top of the screen very quickly, even if my cursor is initially at the bottom. If you're always working in small windows on a very large screen, menu-in-screen may work better for you.
Of course, the problem is that all that research was done assuming small screen sizes and a single-tasking operating system. Today, with a nice wide-screen display, I can have two separate applications up, side by side. I find it highly counter-intuitive that the application on the right will have its menus at the top of the screen, above the other app! Not to mention, it is easy to become confused about which application's menu that is up there, at this particular second.
I used to have an '87 Chevy that had a speed limiter that kicked in right around 115 mph. The engine was running fine, and right at that magic threshold it started sputtering and the speed fell off a bit, and suddenly it was fine again. And no, it was nowhere near redline.
Currently I have an Eagle Talon, which reportedly has a limiter at 145, but I have never put that to the test! If I hadn't read about it, I would have never known.
The justification I remember reading for these speed limiters was that it prevented the car from going faster than the tires were rated for. I don't know if that is really true, though.
Feel free to correct my math if you want. I just don't feel like doing the actual work.
Your problem isn't just your math, it is your understanding of the whole situation. 2 G's is unlikely to cause much harm, but in your scenario, you would experience a whole lot more.
As you step off the building, you fall 32 feet, and in that distance accelerate from 0 to 64 feet-per-second in 2 seconds. Then you hit the ground. Now ask yourself, over what distance do you decelerate from 64 fps to zero? And in how much time?
The results will depend on your body position. If you land upright, you head has a lot more distance to decelerate than your ankles. If you land flat on your back, your whole body decelerates in only a few inches (depending on compressibility) and would experience more than 200 Gs. That will break bones, and most likely kill you.
Where I live, the power company has a similar plan where you can sign up for "green power" from wind and solar installations. This costs an extra 1.5 cents per kWh, and they estimate that to be about $7 per month for the average household. (And that tracks pretty well with my own power bill.)
And yes, they freely admit that it all mixes together in the power grid. But they buy power from the "green" providers based on the usage of the customers signed up for the program. This means they have to burn less fossil fuel to run their own generators.
If you are curious, you can look at their program information here
Anyway, I've never recorded anything audio or data on anything but a data cd. Why would anyone get a special 'audio' one
The audio blanks are needed for the stand-alone CD recorders, i.e. something you would hook into your stereo system like a cassette deck. The audio blanks have a special encoded track at the beginning, in a non-writable area, that these recorders look for. This is all part of some deal worked out years ago for the benefit of the recording industry. Even in the US, these blanks carry an extra surcharge that goes to "repay" the copyright holders. Apparently this deal was worked out before they realized the almost no one uses these devices.
Personally, I know of only one person that actually has one of these. And I still don't understand why he bought it. It seems like you lose a lot of control using one of these, don't have much of any editting capability, and the blanks cost more. So why?
If it currently the warmest it's been in 400 years (or the past few millenia) that means it was this warm 400 years (or a few mllenia) ago.
Sorry, wrong, that is not what they were saying at all. The report states that they have direct temperature records for the past 150 years, very good indirect evidence of temperature for the last 400 years, and weaker indirect evidence for the past 2000 years. While there were fluctuations, including a warm period in the middle ages, there is no indication that in any of those times past it was ever warmer than now.
The National Acadamies summary is better than the Yahoo article, and links to the (155 page!) report itself.
I couldn't find the study itself either, but the news release from the Yankee Group that did the study (here) also states:
In a head-to-head comparison, Windows Server 2003 shows the highest reliability gains, leading Red Hat Enterprise Linux with nearly 20% more annual uptime in similar deployment scenarios.
Now, maybe what they meant was that Windows gained 20% more than Linux gained, which would be meaningless without hard numbers anyhow, but that is not what they said.
If you take their "20% more annual uptime" literally, that would imply that if the Windows box had 100% uptime, the Linux box would have been down for more than 60 days. Clearly, this is gibberish.
No, the idea that XML-based documents AREN'T "inherently" slow is silly. Of course an XML-based document will be slower than a binary document.
You are missing the point. No one is disputing that XML formats are slower than binary formats. But that is a relative comparison, and does not mean XML is slow on an absolute scale. A Porsche may be slower than a Ferrari, but that does not mean a Porsche is slow. (And yes, obviously, this will depend on which Porsche, and which Ferrari, but work with me here...)
And, as you correctly pointed out, there are many advantages of XML over binary formats, which for most of us massively out-weigh the trivial speed difference.
PIs is not an intuitive acronym. And unless yer name is Magnum... don't use it!
The acronym is not the problem so much as the font. Like the original poster, I also read that at P-L-S, and count not guess what it was supposed to mean. The font used for the title of the article makes capital-I and lowercase-L identical.
3) If the monitor is not at the right height (Your eyes should be level with the top edge of the viewable area when you are looking at it with your head tilted slightly down) either adjust the stand or place something STURDY and flat underneath the monitor to raise it to the proper height.
The canonical monitor boosting platform is a few reams of paper. You may need two side by side to be wide enough, depending on your monitor base. You can stack them two high, if you need that much height. While they are still wrapped, they are essentially a solid block of wood, and will easily support the weight of even monster CRTs. They are cheap, easily obtained in most offices, and such time as you are done with them, just take them back to the printer room.
And Angband, of course, was a rip-off of rogue, which is about 10 years older. In fact, hack, nethack, moria, angband, etc are collectively referred to as "roguelikes".
Just like a citizen who is not guilty of any crimes should welcome the friendly government agent asking for his papers...
I have a problem with the "friendly government" requiring me to have proper identifcation papers on my person at all times, so that their agent may ask me for them at his convenience. It is not a matter of guilt or innocence, but one of a budding police state.
No, it isn't. This is a reference to the pilot, where during the battle of Serenity Valley, Mal explains to one of his men that they are not going to die, because they are all too pretty. Go watch it again.
the last time I wrote self modifying code for a legitimate purpose was on the C64
I think most architectures now are not guaranteed to maintain cache coherency. I used to write self-modifying code for 3d stuff on a 486's... it seemed to work then, but by all rights shouldn't have!
Newer architectures do not tend to guarantee cache coherency. However, if there is no hardware cache coherency, then there must be a cache flush instruction. It is needed.
While we don't tend to think of it that way, dynamic library linking is an example of modifying code on the fly. The linker has to overwrite the jump-table in the binary with the locations of the libraries. Then the modified instructions have to be flushed out of the data cache before the code executes, or it might get the old unmodified version and crash.
The thing that scares me about this Intel proposal is that, like "treacherous computing", they are again deciding what may and may not be run on MY computer. Even without any sort of nefarious agenda on their part, I doubt their ability to foresee all possible future legitimate applications that might trip their magic rootkit detector.
When the center receives your call, the operator will not have your address and may not have your phone number on hand, so you must provide that information in order to get help.
Then lower down, it talks about "E911 coming soon", which is what the orginal article was talking about.
Slashdot Mirror
The correct question is "how likely was it for any server to have crashed", and I think the answer is that it was fairly likely.
It is unclear if being New Year's made a crash more likely, but I think we can agree it made it far more likely that we would hear about it.
You didn't pick a random server, you picked one already known to have crashed. And you didn't pick a random 20 second interval either. The odds of that server crashing in that 20 second interval was 100%, because it was already known to have happened. This is a classic mis-application of statistics.
Admittedly, the interval right at New Year's is a bit suspicious, since there is some specific code to handle leap years, etc. But given that there wasn't a rash of outages reported, I am going along with the coincidence theory.
Seriously, what kind of ASSHOLE CHAUVINISTIC PIG would say that the person alleged of sexual assault should it not be true would be WORSE off than the victim were it true?
Read it again. He didn't say that at all. He said that the embarrassment of being publicly identified as a victim of sexual assault is less than the embarrassment of being publicly accused of being the assaulter. He was comparing the results of publication, not of the crime itself.
Actually, OTP does use a simple XOR cipher. Given a truly random key the same length as the message, that is perfectly adequate.
By that logic, a verbatim copy of the whole series would be "factual information" about the contents of the series. This is not like trying to copyright the population of the Congo, or trigonometric tables. This is an original work of fiction, and has no independent existence outside of her work. Based on the description of the lexicon, calling it fair use seems something of a stretch, and calling it "factual information" is a complete farce.
Look again, at the bottom of the page. It is small, but it is there.
Also, I have often heard that a favorite tactic of the Muzak salesmen is to completely ignore this, and try to convince shop owners they find playing the radio that they are violating copyright, and try to get them to sign up to avoid liability. Apparently, this constitutes fraud, but they are rarely called on it. I don't know if the BMI/ASCAP people do the same sorts of things or not.
I hope this case gets slapped down hard by the UK court system, this is pretty ridiculous.
Here it is, from page 4:
This is the last issue of Sys Admin magazine that you will receive. The magazine is ceasing publication as of this issue.The is often a large measure of regret and sadness when a long-term relations ends, and I feel these emotions now at the end of my relationship with Sys Admin. No other publication really does what Sys Admin does, but you probably already know that. You might not know, however, the challenges currently facing many print publications, particularly small niche ones like Sys Admin. These challenges, which have contributed to the decision to stop publishing the magazine, include circulation woes, online competition, and market shift. I take some small comfort in the fact that Sys Admin fought these battles and survived much longer than many others in similar circumstances.
I have worked for Sys Admin magazine for almost 12 years, and I've had a great time. So, in this note, I need to mention some of the many people who have made working for Sys Admin such a rewarding experience for me. In no particular order, I thank Hal Pomeranze, Joe Casad, Ralph Barker, and Robert Ward for making me seem smarter than I really am. I thank Rikki Endsley Kite for therapy shopping, making me do things that scare me, and being a fabulous friend. I thank Lori White and Twyla Watson Bogaard for always reminding me that life goes down better with humor. I thank the regular columnists and writers for their loyalty and consistently excellent articles, and I thank all the readers and contributors for making Sys Admin as successful as it was for as long as it was.
Good-bye and good luck.
Sincerely yours,
Amber Ankerholz
Editor in Chief
Of course, the problem is that all that research was done assuming small screen sizes and a single-tasking operating system. Today, with a nice wide-screen display, I can have two separate applications up, side by side. I find it highly counter-intuitive that the application on the right will have its menus at the top of the screen, above the other app! Not to mention, it is easy to become confused about which application's menu that is up there, at this particular second.
Currently I have an Eagle Talon, which reportedly has a limiter at 145, but I have never put that to the test! If I hadn't read about it, I would have never known.
The justification I remember reading for these speed limiters was that it prevented the car from going faster than the tires were rated for. I don't know if that is really true, though.
Your problem isn't just your math, it is your understanding of the whole situation. 2 G's is unlikely to cause much harm, but in your scenario, you would experience a whole lot more.
As you step off the building, you fall 32 feet, and in that distance accelerate from 0 to 64 feet-per-second in 2 seconds. Then you hit the ground. Now ask yourself, over what distance do you decelerate from 64 fps to zero? And in how much time?
The results will depend on your body position. If you land upright, you head has a lot more distance to decelerate than your ankles. If you land flat on your back, your whole body decelerates in only a few inches (depending on compressibility) and would experience more than 200 Gs. That will break bones, and most likely kill you.
And yes, they freely admit that it all mixes together in the power grid. But they buy power from the "green" providers based on the usage of the customers signed up for the program. This means they have to burn less fossil fuel to run their own generators.
If you are curious, you can look at their program information here
The audio blanks are needed for the stand-alone CD recorders, i.e. something you would hook into your stereo system like a cassette deck. The audio blanks have a special encoded track at the beginning, in a non-writable area, that these recorders look for. This is all part of some deal worked out years ago for the benefit of the recording industry. Even in the US, these blanks carry an extra surcharge that goes to "repay" the copyright holders. Apparently this deal was worked out before they realized the almost no one uses these devices.
Personally, I know of only one person that actually has one of these. And I still don't understand why he bought it. It seems like you lose a lot of control using one of these, don't have much of any editting capability, and the blanks cost more. So why?
Sorry, wrong, that is not what they were saying at all. The report states that they have direct temperature records for the past 150 years, very good indirect evidence of temperature for the last 400 years, and weaker indirect evidence for the past 2000 years. While there were fluctuations, including a warm period in the middle ages, there is no indication that in any of those times past it was ever warmer than now.
The National Acadamies summary is better than the Yahoo article, and links to the (155 page!) report itself.
Now, maybe what they meant was that Windows gained 20% more than Linux gained, which would be meaningless without hard numbers anyhow, but that is not what they said.
If you take their "20% more annual uptime" literally, that would imply that if the Windows box had 100% uptime, the Linux box would have been down for more than 60 days. Clearly, this is gibberish.
You are missing the point. No one is disputing that XML formats are slower than binary formats. But that is a relative comparison, and does not mean XML is slow on an absolute scale. A Porsche may be slower than a Ferrari, but that does not mean a Porsche is slow. (And yes, obviously, this will depend on which Porsche, and which Ferrari, but work with me here...)
And, as you correctly pointed out, there are many advantages of XML over binary formats, which for most of us massively out-weigh the trivial speed difference.
Therefore, unions create larger markets.
Correlation is not causation.
The acronym is not the problem so much as the font. Like the original poster, I also read that at P-L-S, and count not guess what it was supposed to mean. The font used for the title of the article makes capital-I and lowercase-L identical.
The canonical monitor boosting platform is a few reams of paper. You may need two side by side to be wide enough, depending on your monitor base. You can stack them two high, if you need that much height. While they are still wrapped, they are essentially a solid block of wood, and will easily support the weight of even monster CRTs. They are cheap, easily obtained in most offices, and such time as you are done with them, just take them back to the printer room.
You might take a look at Wikipedia's article
I have a problem with the "friendly government" requiring me to have proper identifcation papers on my person at all times, so that their agent may ask me for them at his convenience. It is not a matter of guilt or innocence, but one of a budding police state.
No, it isn't. This is a reference to the pilot, where during the battle of Serenity Valley, Mal explains to one of his men that they are not going to die, because they are all too pretty. Go watch it again.
I think most architectures now are not guaranteed to maintain cache coherency. I used to write self-modifying code for 3d stuff on a 486's... it seemed to work then, but by all rights shouldn't have!
Newer architectures do not tend to guarantee cache coherency. However, if there is no hardware cache coherency, then there must be a cache flush instruction. It is needed.
While we don't tend to think of it that way, dynamic library linking is an example of modifying code on the fly. The linker has to overwrite the jump-table in the binary with the locations of the libraries. Then the modified instructions have to be flushed out of the data cache before the code executes, or it might get the old unmodified version and crash.
The thing that scares me about this Intel proposal is that, like "treacherous computing", they are again deciding what may and may not be run on MY computer. Even without any sort of nefarious agenda on their part, I doubt their ability to foresee all possible future legitimate applications that might trip their magic rootkit detector.
The page states that they have 911 service, but:
Then lower down, it talks about "E911 coming soon", which is what the orginal article was talking about.
If you RTFA, you will see it is Japanese capacitors this time, instead of Taiwanese, but basically the same problem as three years ago.