Just looked at libdispatch source code. It is apparently very Mac OS X specific, and will take a lot of work to be ported cross platform. The queue implementation also looks like it imposes a lot of overhead, so it is not very useful for parallelizing short-running "blocks" of code.
Why not use something like Cilk which has proven its salt? The only caveat of Cilk is that it requires any parallel code to be compiled with the Cilk compiler (which will generate C code), and unless you're willing to kickstart the Cilk runtime manually, any caller of the parallel code needs to be compiled with Cilk too, which includes main(). The technical reason is Cilk does a continuation passing style transformation (requiring a different calling convention) which allows a function caller to be stolen by the work-stealing scheduler while the function waits for the callee to finish, allowing the critical path to focus on work instead of manipulating the task queue. The result is a highly efficient and scalable parallel execution runtime, the best I've seen.
Reliant Technology sells you NetApp FAS 6040 for $78,500 with a maximum capacity of 840 drives, without the hard drive (source: Google Shopping). If you buy FAS 6040 with the drives, most vendors will use more expensive and less capacity 15k rpm drives instead of the 7200rpm drives the BlackBlaze Pod uses, and this makes up a lot of the price difference. The point is, you could buy NetApp and install it yourself with cheap off-the-shelf consumer drives and end up spending about the same magnitude amount of money. I estimate that NetApp would cost just 1.5x the amount.
NetApp FAS 6040 at $78,500 + 840 x 1.5TB drives at $120 each = $179,300 which gives you 1.26PB. Cost per petabyte is $142,500, only slightly more expensive than BlackBlaze $117,000 from the article. The real story is that BlackBlaze is able to show a competitive edge of $30,000, or being 20% cheaper.
If you use a software tracking approach, thieves will learn to put tape over the webcam, and not use it until they ask someone in the black market to reinstall the operating system for a fee, or learn to reinstall operating system by themselves. This is just like how thieves learned to use the credit card immediately after they steal it before you have a chance to cancel the card.
Other approaches taken will lead to workarounds too. If you have a database of serial numbers and require legit second hand market to verify the serial number before going on sale, the thief will just keep the gadget for themselves, for their friends, or sell it in the black market. If you're able to remotely brick the device, the thief will still be able to use it for a while, and then just steal another one. I'm sure thieves don't mind using new gadgets all the times.
Consider that they were the black hats forefathers who inspired the hackers in Russian and China, I think their limit is not whether they can do it or not, but it's what they end up doing that matters.
So a nerve circuit that controls how you enunciate is ridiculously complicated in humans but simple in fish. As if a fish talks. Show me a talking tilapia, then we can talk.
Oh, found the code on lxr. It looks like Linux kernels up to 2.6.29.6 are NOT affected, and this is a vulnerability introduced in 2.6.30 due to a fairly significant rewrite of tun.c. Linux 2.6.30 was released in Jun 9, 2009, just a month ago. Funny the tun.c rewrite was not mentioned in the set of changes for 2.6.30.
I think this example actually shows a forte of Linux as open source. New vulnerability is found very quickly after "new" code is released.
I tried to google code search for "tun->sk" and Linux doesn't contain that snippet of code. Since SANS claimed that drivers/net/tun.c is at fault, I looked at that source file and didn't find any instances where "if (!...) return...;" is performed after NULL dereference.
I think the only fascinating bit of the story is that the SElinux extension allows you to map a page at memory address 0 (the NULL page), making NULL dereferencing valid. I also found out about that a while ago, but I didn't know it has anything to do with SElinux. By the way, mapping the NULL page also works on Mac OS X.
However, mapping NULL page is typically NOT exploitable. A correct program will simply reject access to NULL pointer, giving it a special semantic regardless whether the memory page itself is valid or not.
Read Matthew chapter 27 and 28 from Codex Sinaiticus. These two chapters describe the account of Jesus' death and resurrection, and there is no omission from the Codex. Amen.
Rather than advocating a specific anti-virus product, I feel that the question is how do you know you need anti-virus. I would recommend choosing two anti-virus product and keep them up to date for the shortest subscription period allowed. During this time, work as usual, and take note of any virus alert you get, and how that happens. Get rid of one of the anti-virus that doesn't appear to be as effective (and recommend the remaining one to other people). Also adjust your computer using habit until you get no virus alerts. Then make sure you keep your habit within the confines of rules you find working well for you, so you don't get virus alert. Then get rid of the anti-virus software altogether.
I regard anti-virus software as some sort of potty training. You only need it until you find out what behavior will get you into trouble.
Walk into a bookstore, look at some magazines about audio and sound engineering, and contact the publisher to see if they're interested in reviewing your software. I can recall one prominent magazine called Sound on Sound, but you should be able to find more.
What puzzles me is that, intuitively, the cost of infrastructure construction is amortized to be negligibly small over many runs of freight. However, the report still claims that it is a significant factor. Also, data are shown only up to regional commuter rail, but high speed trains are not compared in this report. On the other hand, they show data for large commercial aircrafts.
No it's not. If you ask someone to do you a favor and expect that person to comply, that is not a favor; that is bullying. The other person has all the rights to refuse your request. Not accepting "no" as a response is rude and anti-social. (As an aside, I almost never give money to homeless because they often have this attitude that I owe them money and everything I have.)
In the case of this story submitter, I think the most difficult psychological barrier is to tell people the real reason he doesn't want to lend his laptop: "I don't feel comfortable letting anyone use my laptop." It's okay to be a bit selfish. People who regard you as a friend will respect that and keep you a friend.
If you are so compelled to help those who don't have a computer to check e-mail or browse the web, then that's the proper context under which discussing technical methods to protect the privacy of the computer could be helpful. In which case, guest account, virtual machine, Google Chrome incognito mode are all fairly good ways.
The question now would be, how do you comfortably ask them to yield the laptop when you do need to use it. That comes back to the point that, if you're doing someone a favor, they should be able to take no for an answer, even ever so temporarily.
If someone sees that corporal punishment helped in their life, of course it logically follows that they would do the same to their children to help them the same. Those who see this undesirable are the outsiders who do not receive corporal punishment and thus do not enjoy the benefits of it.
How did corporal punishment start in a family? Possibly at one point, a parent became frustrated with his own problems and wanted his children to live a different life. Such person might not know a better way to discipline except by physical means. If you know a better way to discipline, feel free to use it. However, it is no denying that discipline will lead to success and a better life. Telling a parent that he is not allowed to discipline his children is just plain arrogant. Enforcing that systematically, you get a civilization of sloths, which is what America is becoming to.
Japanese characters (the real Japanese ones, not the ones borrowed from Chinese) just seem so beautiful to me
Hiraganas are rooted in Chinese cursive form of the characters.
And also, I've become convinced that the best way to kill your romantic vision of a place is to live there.
The best way to want living in a place even though you absolutely hate it is to have strong ties to local people there. This is easier said than done though.
To give him fair credit, Slashdot posted a study back in February 2006 How Songs Get Popular. Here is a quote from TFA:
Researchers created an artificial "music market" of 14,341 participants drawn from a teen-interest Web site. Upon entering the study's Internet market, the participants were randomly, and unknowingly, assigned to either an "independent" group or a "social influence" group.... Researchers found that popular songs were popular and unpopular songs were unpopular, regardless of their quality established by the other group. They also found that as a particular songs' popularity increased, participants selected it more often.... Although different songs were hits in each world, popularity was still the deciding factor, although the "best" songs never did very badly and the "worst" songs never did very well.
I think this is still consistant with the submitter's claims. Calling him delusional just because he represents an indie label is a personal attack.
That was a very poor argument. You're basing your argument on the top 100 torrents, this is like an inverse of the "long tail" argument. But that's the only data you have, since you can't look at the top 100,000 torrents.
Even if you look at the top 100,000 torrents, you'd still find that most of it is going to be releases from commercial artists. Not that there are 100,000 artists, but most torrents share one album at a time (each artist has multiple album releases), and there are lots of dups.
A more realistic measure is the seeder and leecher count. Most of the commercial artists have thousands of seeders and leechers at any given time. Some of the less popular stuff I tried to download have 0 seeder and 1 leecher; that leecher was me. This should give you an idea of the unfairness of the distribution.
I truthfully don't care about music. What I care about is when textbooks start becoming free. It will be a revolution in education. This will be especially the case when people write things like,"The comprehensive guide to calculus as to be learned by anyone who knows how to count" The computer means it can be an advanced and interactive media session. The free distribution will mean anyone can have it in their hands.
If it helps giving you a sense why you'd very unlikely get free and quality textbooks: I know enough calculus that I'm quite willing to teach you for free when I have nothing else to do, but writing a textbook takes an organized effort beyond what I'm willing to do for free.
Not to say that writing a textbook about emerging fields for graduate students (or advanced undergraduate students) involves hours and hours of survey and comprehensive study of related papers. These papers only describe incremental developments of a concept, and are targeted only for researchers in the same field. The papers are, in a sense, a bush in the forest, and writing a textbook requires you to map out the entire forest by looking at a bush at a time. Then you need to figure out how to organize it in a way so the material is accessible to first-time readers of that subject.
These messenger boys (don't know about typists) were probably there because they come from a poor family and didn't have the means of proper education. However, they could learn much on the job by interacting with and observing the professionals. Some of the brightest who are willing to learn on their own could actually gain a successful career one day because of the experience they gathered doing these low-skill service positions. I'm sure you can find many autobiographies of successful people who began their lives similarly.
Nowadays they are replaced by automation. That means the poor and uneducated lose a valuable opportunity to become successful. Their only chance now is to go through a proper education, and our education system still favors in many ways families living comfortable lives.
Many anti-virus software spun up the floopy drive when Windows 95 was starting up, and when it was shutting down. The way I see it, Windows 95 could do a read with or without the floppy at start up. If an error is returned, it would indicate whether there is a disk or not or if that's a different kind of error. Compare the result with the algorithm that detects presence of a disk without spinning up. Just need to do this once during boot time. Don't even bother caching the result in the registry since a user could have replaced a floppy drive with an "opposite" one.
Another idea is to always show the floopy disk icon when the system first starts, regardless whether there is a disk in the drive. If the user clicks on the icon, attempts a read. Determine the polarity of the spinless floppy presence algorithm at this point and compare the result. If read succeeds, keep the icon. If the read fails, ask the user to insert a disk to try again or cancel. If the user cancels, remove the icon. You have now determined the polarity and can do spinless detection reliably.
I think the summary writer means UNIX servers actually used by businesses to store data and conduct transactions. I don't think Xserve is quite there yet although it is an impressive piece of work.
My friend who studied communications has a Bachelor of Science degree, and I who studied computer science have a Bachelor of Arts degree. A quick google showed that both are fairly common.
Interesting experience how to cope with the stupidity of Windows. We'd all keep that in mind if we end up administering a network of Windows machines. But on the Linux side, things aren't as complicated.
Anything they need to do their job will already be installed and if it's not, we need to ensure we have a proper license before allowing it on the systems.
On Linux, people will have difficulty finding unlicensed software since most are freely licensed under one of the open source licenses.
Next, I have seen things go seriously wrong when it never should have because of users installing their own apps.... The installation replaced a few codec and defaulted some extensions to the new player and all the sudden we didn't have a machine on the floor that could view the security video footage.
On Linux, you install things under your home directory. The installation is only seen by the user, and the system-wide software is not affected. You login as another user on the system and you only see the standard software. This holds as long as you don't give users the permission to install.rpm or.deb. The user can always use 0install, autopackage, or old fashioned configure && make && make install.
I have also seen people install apps to view their damn E-greeting cards that not only resulted in Email addresses for the entire firms being harvested and sold, but for every client that had ever used an email address with the company
That is a tough one. Leaking information isn't something that needs admin privilege to do. I don't think Linux is immune to that either.
Maybe someone else has more insight on this one?
I had one person install one of those remote desktop programs so he can access shit from home. Well, his home computer was a virus riddled clusterfuck that not only proceeded to place some nasty shit on the network, but allowed someone completely unknown who got the logon from the Trojans on the system and had access to the system before we got it caught.
I think the reason why he got virus from home to the work machine is because he shared local disk on the work machine to the remote home machine. On Linux, "rdesktop" also allows you to do that. It is akin to someone bringing his USB thumb drive back and forth from work to home. Do you block USB drives too?
I suppose you could compile "rdesktop" from source and remove the disk share feature before you compile. After all, you have the source code, and you do whatever you want with it.
Also, most Windows viruses won't work on Linux, thankfully.
Finally, VNC doesn't seem to support file sharing feature, so that should be safe as a remote desktop viewer.
In office 2000 there was an issue with service pack 2 (it may have been SP1 but two jumps to mind) where if you cut and pasted from multiple sources and used hyperlinks, it would fill the clipboard up and instead of releasing the memory, it would then fill that up until the program crashed and you lost everything (auto-save didn't seem to work when it crashed like that, you had to manually hit save every so often).
Why not make it auto-save every minute, or every time the user stops typing? Auto-save every two hours seems to be too long.
I hope you personally tested OpenOffice rigorously because I haven't, and I can't voucher for the quality of that software, whether it has memory leaks or something like that. I just know that Sun has a cathedral style of managing the development of OpenOffice, and it's hurting the developer community.
Slashdot Mirror
Just looked at libdispatch source code. It is apparently very Mac OS X specific, and will take a lot of work to be ported cross platform. The queue implementation also looks like it imposes a lot of overhead, so it is not very useful for parallelizing short-running "blocks" of code.
Why not use something like Cilk which has proven its salt? The only caveat of Cilk is that it requires any parallel code to be compiled with the Cilk compiler (which will generate C code), and unless you're willing to kickstart the Cilk runtime manually, any caller of the parallel code needs to be compiled with Cilk too, which includes main(). The technical reason is Cilk does a continuation passing style transformation (requiring a different calling convention) which allows a function caller to be stolen by the work-stealing scheduler while the function waits for the callee to finish, allowing the critical path to focus on work instead of manipulating the task queue. The result is a highly efficient and scalable parallel execution runtime, the best I've seen.
Let human provide the motivation. Oh, wait...
Reliant Technology sells you NetApp FAS 6040 for $78,500 with a maximum capacity of 840 drives, without the hard drive (source: Google Shopping). If you buy FAS 6040 with the drives, most vendors will use more expensive and less capacity 15k rpm drives instead of the 7200rpm drives the BlackBlaze Pod uses, and this makes up a lot of the price difference. The point is, you could buy NetApp and install it yourself with cheap off-the-shelf consumer drives and end up spending about the same magnitude amount of money. I estimate that NetApp would cost just 1.5x the amount.
NetApp FAS 6040 at $78,500 + 840 x 1.5TB drives at $120 each = $179,300 which gives you 1.26PB. Cost per petabyte is $142,500, only slightly more expensive than BlackBlaze $117,000 from the article. The real story is that BlackBlaze is able to show a competitive edge of $30,000, or being 20% cheaper.
If you use a software tracking approach, thieves will learn to put tape over the webcam, and not use it until they ask someone in the black market to reinstall the operating system for a fee, or learn to reinstall operating system by themselves. This is just like how thieves learned to use the credit card immediately after they steal it before you have a chance to cancel the card.
Other approaches taken will lead to workarounds too. If you have a database of serial numbers and require legit second hand market to verify the serial number before going on sale, the thief will just keep the gadget for themselves, for their friends, or sell it in the black market. If you're able to remotely brick the device, the thief will still be able to use it for a while, and then just steal another one. I'm sure thieves don't mind using new gadgets all the times.
Consider that they were the black hats forefathers who inspired the hackers in Russian and China, I think their limit is not whether they can do it or not, but it's what they end up doing that matters.
So a nerve circuit that controls how you enunciate is ridiculously complicated in humans but simple in fish. As if a fish talks. Show me a talking tilapia, then we can talk.
For some reason I didn't link this correctly. The set of changes for 2.6.30 is found http://kernelnewbies.org/Linux_2_6_30.
Oh, found the code on lxr. It looks like Linux kernels up to 2.6.29.6 are NOT affected, and this is a vulnerability introduced in 2.6.30 due to a fairly significant rewrite of tun.c. Linux 2.6.30 was released in Jun 9, 2009, just a month ago. Funny the tun.c rewrite was not mentioned in the set of changes for 2.6.30.
I think this example actually shows a forte of Linux as open source. New vulnerability is found very quickly after "new" code is released.
I tried to google code search for "tun->sk" and Linux doesn't contain that snippet of code. Since SANS claimed that drivers/net/tun.c is at fault, I looked at that source file and didn't find any instances where "if (!...) return ...;" is performed after NULL dereference.
I think the only fascinating bit of the story is that the SElinux extension allows you to map a page at memory address 0 (the NULL page), making NULL dereferencing valid. I also found out about that a while ago, but I didn't know it has anything to do with SElinux. By the way, mapping the NULL page also works on Mac OS X.
However, mapping NULL page is typically NOT exploitable. A correct program will simply reject access to NULL pointer, giving it a special semantic regardless whether the memory page itself is valid or not.
Here is the report
Read Matthew chapter 27 and 28 from Codex Sinaiticus. These two chapters describe the account of Jesus' death and resurrection, and there is no omission from the Codex. Amen.
Rather than advocating a specific anti-virus product, I feel that the question is how do you know you need anti-virus. I would recommend choosing two anti-virus product and keep them up to date for the shortest subscription period allowed. During this time, work as usual, and take note of any virus alert you get, and how that happens. Get rid of one of the anti-virus that doesn't appear to be as effective (and recommend the remaining one to other people). Also adjust your computer using habit until you get no virus alerts. Then make sure you keep your habit within the confines of rules you find working well for you, so you don't get virus alert. Then get rid of the anti-virus software altogether.
I regard anti-virus software as some sort of potty training. You only need it until you find out what behavior will get you into trouble.
Walk into a bookstore, look at some magazines about audio and sound engineering, and contact the publisher to see if they're interested in reviewing your software. I can recall one prominent magazine called Sound on Sound, but you should be able to find more.
What puzzles me is that, intuitively, the cost of infrastructure construction is amortized to be negligibly small over many runs of freight. However, the report still claims that it is a significant factor. Also, data are shown only up to regional commuter rail, but high speed trains are not compared in this report. On the other hand, they show data for large commercial aircrafts.
No it's not. If you ask someone to do you a favor and expect that person to comply, that is not a favor; that is bullying. The other person has all the rights to refuse your request. Not accepting "no" as a response is rude and anti-social. (As an aside, I almost never give money to homeless because they often have this attitude that I owe them money and everything I have.)
In the case of this story submitter, I think the most difficult psychological barrier is to tell people the real reason he doesn't want to lend his laptop: "I don't feel comfortable letting anyone use my laptop." It's okay to be a bit selfish. People who regard you as a friend will respect that and keep you a friend.
If you are so compelled to help those who don't have a computer to check e-mail or browse the web, then that's the proper context under which discussing technical methods to protect the privacy of the computer could be helpful. In which case, guest account, virtual machine, Google Chrome incognito mode are all fairly good ways.
The question now would be, how do you comfortably ask them to yield the laptop when you do need to use it. That comes back to the point that, if you're doing someone a favor, they should be able to take no for an answer, even ever so temporarily.
If someone sees that corporal punishment helped in their life, of course it logically follows that they would do the same to their children to help them the same. Those who see this undesirable are the outsiders who do not receive corporal punishment and thus do not enjoy the benefits of it.
How did corporal punishment start in a family? Possibly at one point, a parent became frustrated with his own problems and wanted his children to live a different life. Such person might not know a better way to discipline except by physical means. If you know a better way to discipline, feel free to use it. However, it is no denying that discipline will lead to success and a better life. Telling a parent that he is not allowed to discipline his children is just plain arrogant. Enforcing that systematically, you get a civilization of sloths, which is what America is becoming to.
Hiraganas are rooted in Chinese cursive form of the characters.
The best way to want living in a place even though you absolutely hate it is to have strong ties to local people there. This is easier said than done though.
To give him fair credit, Slashdot posted a study back in February 2006 How Songs Get Popular. Here is a quote from TFA:
I think this is still consistant with the submitter's claims. Calling him delusional just because he represents an indie label is a personal attack.
Even if you look at the top 100,000 torrents, you'd still find that most of it is going to be releases from commercial artists. Not that there are 100,000 artists, but most torrents share one album at a time (each artist has multiple album releases), and there are lots of dups.
A more realistic measure is the seeder and leecher count. Most of the commercial artists have thousands of seeders and leechers at any given time. Some of the less popular stuff I tried to download have 0 seeder and 1 leecher; that leecher was me. This should give you an idea of the unfairness of the distribution.
If it helps giving you a sense why you'd very unlikely get free and quality textbooks: I know enough calculus that I'm quite willing to teach you for free when I have nothing else to do, but writing a textbook takes an organized effort beyond what I'm willing to do for free.
Not to say that writing a textbook about emerging fields for graduate students (or advanced undergraduate students) involves hours and hours of survey and comprehensive study of related papers. These papers only describe incremental developments of a concept, and are targeted only for researchers in the same field. The papers are, in a sense, a bush in the forest, and writing a textbook requires you to map out the entire forest by looking at a bush at a time. Then you need to figure out how to organize it in a way so the material is accessible to first-time readers of that subject.
These messenger boys (don't know about typists) were probably there because they come from a poor family and didn't have the means of proper education. However, they could learn much on the job by interacting with and observing the professionals. Some of the brightest who are willing to learn on their own could actually gain a successful career one day because of the experience they gathered doing these low-skill service positions. I'm sure you can find many autobiographies of successful people who began their lives similarly.
Nowadays they are replaced by automation. That means the poor and uneducated lose a valuable opportunity to become successful. Their only chance now is to go through a proper education, and our education system still favors in many ways families living comfortable lives.
Many anti-virus software spun up the floopy drive when Windows 95 was starting up, and when it was shutting down. The way I see it, Windows 95 could do a read with or without the floppy at start up. If an error is returned, it would indicate whether there is a disk or not or if that's a different kind of error. Compare the result with the algorithm that detects presence of a disk without spinning up. Just need to do this once during boot time. Don't even bother caching the result in the registry since a user could have replaced a floppy drive with an "opposite" one.
Another idea is to always show the floopy disk icon when the system first starts, regardless whether there is a disk in the drive. If the user clicks on the icon, attempts a read. Determine the polarity of the spinless floppy presence algorithm at this point and compare the result. If read succeeds, keep the icon. If the read fails, ask the user to insert a disk to try again or cancel. If the user cancels, remove the icon. You have now determined the polarity and can do spinless detection reliably.
I think the summary writer means UNIX servers actually used by businesses to store data and conduct transactions. I don't think Xserve is quite there yet although it is an impressive piece of work.
My friend who studied communications has a Bachelor of Science degree, and I who studied computer science have a Bachelor of Arts degree. A quick google showed that both are fairly common.
Interesting experience how to cope with the stupidity of Windows. We'd all keep that in mind if we end up administering a network of Windows machines. But on the Linux side, things aren't as complicated.
On Linux, people will have difficulty finding unlicensed software since most are freely licensed under one of the open source licenses.
On Linux, you install things under your home directory. The installation is only seen by the user, and the system-wide software is not affected. You login as another user on the system and you only see the standard software. This holds as long as you don't give users the permission to install .rpm or .deb. The user can always use 0install, autopackage, or old fashioned configure && make && make install.
That is a tough one. Leaking information isn't something that needs admin privilege to do. I don't think Linux is immune to that either.
Maybe someone else has more insight on this one?
I think the reason why he got virus from home to the work machine is because he shared local disk on the work machine to the remote home machine. On Linux, "rdesktop" also allows you to do that. It is akin to someone bringing his USB thumb drive back and forth from work to home. Do you block USB drives too?
I suppose you could compile "rdesktop" from source and remove the disk share feature before you compile. After all, you have the source code, and you do whatever you want with it.
Also, most Windows viruses won't work on Linux, thankfully.
Finally, VNC doesn't seem to support file sharing feature, so that should be safe as a remote desktop viewer.
Why not make it auto-save every minute, or every time the user stops typing? Auto-save every two hours seems to be too long.
I hope you personally tested OpenOffice rigorously because I haven't, and I can't voucher for the quality of that software, whether it has memory leaks or something like that. I just know that Sun has a cathedral style of managing the development of OpenOffice, and it's hurting the developer community.