Ok, so you want to store data in the unused parts of the filesystem. How are you going to keep track of it? Well, a tried-and-true way is to use the concept of addressing through blocks and then use an inode/block/extent structure to keep track of which data corresponding to which file is in which block. And you know what? You'd *STILL* probably waste space, so why not implement a mapping structure which keeps track of data stored in the holes of the blocks stored in the holes of the blocks of the original filesystem.
In other words...you'd have to implement a filesystem again, using smaller blocks, and gain nothing. Worse than nothing, since the "first" filesystem will be keeping track of its own data in the first part of each block, the "second" filesystem will have to cope with variable-sized blocks, which means more overhead.
If you want to waste less space on your drive (silly, considering the cost per gig these days) just use smaller blocks. And suffer a performance hit, of course, and reduce the upper limit on the size of your drive, and...
The only dynamic linking the kernel uses is modules, which aren't used for providing library routines like zlib. The kernel does not link.so files. The code is almost certainly cut-and-pasted into the ppp compression code somewhere.
Wrong, too. The aspect ratio of the movie screen depends on the film format. 35mm is 16x9, but 70mm and 135mm are much wider. Most movies today are indeed shot in 35mm, but not all, particularly epic movies (Last of the Mohicans comes to mind).
So in Plot Originality, Visual Impression, and Film Transfer, the DVD is better than 995 out of 1000 other DVDs (at least, I suppose his scale must be logarithmic or exponential?)
I was one of the Dell laptop users whose batteries had a chance of catching on fire. Under Dell's replacement plan, they sent me a new battery immediately, and whenever I send back the defective one I get a second battery. That's a hell of a deal... I hardly use my batteries, it's mostly a desktop replacement, but I'm worried about the batteries dying completely someday and making the laptop plug-in only. I'm planning on saving the second battery until the first one dies. Thanks, Dell!
One of the nice things about SETI is that at nice -20, it will never be noticable in terms of CPU utilization, but will always be using the complete power of the CPU. Could we do that with disks?
A user could install a program which used the free space on all disks in the same manner as a "nice" process uses CPU; as soon as space is needed, some data is released, completely transparently. A company or organization could store data on the distributed network; they would keep a "master" copy of the data available, in case a particular fragment happened to be erased on all of the nodes, or nodes were unavailable.
The question I'm pondering is how to keep track of where data is stored, and route data from the nodes to the host where it would be read. In article's example, the fragments of a movie, sent to a particular client. How do we efficiently request fragments, in the correct order, without either overusing bandwidth with duplicated data or dropping fragments?
The funny thing is that they're implying that the ReplayTV customer can "steal" or "magically acquire" those X-Files episodes or James Bond movies...the very same episodes and movies which the networks are broadcasting via very powerful transmitters. Gee, if they were so worried about people stealing their content, you think they wouldn't give it away...
Fox can easily prevent X-Files watchers from acquiring copies of the episodes. Just don't broadcast them.
The good thing is that in courts, the argument of "if they do this it will hurt our business" doesn't hold up, even for baseball and it's strange exemption from antitrust laws.
Huh? Active Directory (and Exchange before that) is nothing but LDAP. DOM and HTML/HTTP are both obviously in IE. Kerberos is supported, certainly IMAP...
Microsoft has in fact been *catching up* to these standards for a few years now. Doesn't that sound interesting?
...it's a win for Intel. Larry says nothing in the article about the capabilities of Linux except that it's better than Windows "if you're on the Internet."
What he really liked, apparently, was the fact that the hardware was cheap and easily replaceable. It's a win for clustering, certainly, but is it a win for Linux?
He was always a fan of interesting numbers, people who had addresses like 1234 Main Street. For a long time he wanted a phone number with all ten (or at least seven) digits the same, but couldn't get one. Finally when the 888 toll-free area code came out, he was able to get a cell phone with the number (888)888-8888.
Soon thereafter, he began getting mysterious calls on his new phone. The phone would ring and there would be just silence, or strange (but not particularly rude) noises. These happened several times a day. Eventually, in one of those calls, he heard a woman's voice: "What are you doing with that? Put that down." followed by the other end hanging up.
He figured out that it was babies who were calling him. If a baby or young child picks up the phone, one of the most likely numbers to dial is the same digit over and over. Kids were picking up the phone and mashing the 8 button constantly.
I read this in a Wired interview (doesn't seem to be online) which ended with the line, "...the babies of the world were calling the Woz."
You can't get your Jack Bot pregnant, so there's no need to order a new Rubber.
Re:What came first, the GNU kernel or linux?
on
Debian NetBSD
·
· Score: 3, Informative
Well, no, it was Linus himself who first paired the kernel with GNU tools. He developed it using GNU development tools on Minix and then Linux itself once it was usable, and ASAP he had the standard GNU toolset compiled and running on his own box.
Perhaps Slackware did indeed release the first distribution; but from the start, GNU was used on Linux. I imagine that the 0.1 release had some mention of "BTW, if you want to actually DO anything, go get the GNU stuff..."
For one thing it's a quarter million, not a quarter billion, and for another I'm never going to be impressed by a number of lines of code, but by how well it works.
What do you mean, it wasn't literal? Clarke and Kubrick obviously thought about things they thought would be happening in the near future. I seem to recall Clarke being pessimistic about an AI as smart as HAL, but that's not quite enough to label the date of 2001 as "not literal." In the book, the events clearly happen in the year 2001 AD (or most of them, anyway). 2001 is much more specific and literal than a dystopian book like 1984 (where I would agree the date is more symbolic).
Science fiction is never completely accurate, obviously. But Clarke was one of the most accurate and scientifically rational writers of the century. We haven't gotten to convenient interplanetary travel quite yet, but you can be sure that it will happen much like he describes: a large space station using 'centrifigal force' to simulate gravity, and rockets using the station as a waypoint so the same spacecraft doesn't have to be capable of lifting off from Earth as well as travelling to and landing on another planet or moon.
Now, being able to phone from the station to America for only a few dollars, that's probably a little over-optimistic...
NT is certainly nothing like what it was originally intended to be. I've heard that it was developed on Alphas and then ported to the i386, just to make sure the developers made portability a priority. Look how long THAT lasted...
However, NT is POSIX compliant, just as much as Linux is, in fact. POSIX is a very general and practically useless standard; it's very easy to implement because it defines very little, and leaves many important considerations out.
All data is NOT WRITTEN TWICE. RAID is a solution for replicating data. Journalling file systems replicate meta-data, which is the information ABOUT a file, such as its name and where it's stored on disk. They eliminate the need for fsck, which will not recover lost data either. Before you bash it, understand it.
If I were writing an application which had a hope of running remotely (a standard windowed word processor for example) I would write it to X. But if I were writing a new flight simulator, I would know up front that there is no hope of running it remotely, because it needs direct hardware acceleration, and I would write it for the DirectFB layer.
This is more like DirectX than anything; a way to bypass the high-level windowing system to write directly to hardware. As people have said, it doesn't replace X completely. But I would rather have a X server layer on top of a direct-hardware layer, than a direct hardware piece hacked into an old X server.
The problem with DRM is the same as with computer security. It's very easy to secure a computer so that no one can use it. It's very difficult to secure a computer that a great many people need to use in a great many ways, while at the same time restricting any unauthorized users whatsoever.
Media companies really want to stop unauthorized use of their copyrighted material. Copying a CD and giving it to someone else is illegal. Copying a CD or creating a compilation for your own personal use is legitimate. The problem is allowing the legal copies while preventing the illegal ones. And it's a very difficult one to solve. It's much easier to simply prevent all copies whatsoever.
Anyone who thinks that "fair use" means giving away copies of music or books is a thief and an idiot. Remember this: if you stop paying for your media, they will stop selling it to you.
Slashdot Mirror
Oh, yeah, you know, you can fix PROBLEM_X with one command line:
/usr/x22/lib/bin fooprog -command"somethingorother" ls02 sk202 -a -w 3-2 e-d -s -ddd- -s akseod ddledldl
Just because YOU know how to do it doesn't mean it's easy.
You had to ask female friends to find out that real breasts aren't anything like that? I'm sorry, my friend.
Ok, so you want to store data in the unused parts of the filesystem. How are you going to keep track of it? Well, a tried-and-true way is to use the concept of addressing through blocks and then use an inode/block/extent structure to keep track of which data corresponding to which file is in which block. And you know what? You'd *STILL* probably waste space, so why not implement a mapping structure which keeps track of data stored in the holes of the blocks stored in the holes of the blocks of the original filesystem.
In other words...you'd have to implement a filesystem again, using smaller blocks, and gain nothing. Worse than nothing, since the "first" filesystem will be keeping track of its own data in the first part of each block, the "second" filesystem will have to cope with variable-sized blocks, which means more overhead.
If you want to waste less space on your drive (silly, considering the cost per gig these days) just use smaller blocks. And suffer a performance hit, of course, and reduce the upper limit on the size of your drive, and...
The only dynamic linking the kernel uses is modules, which aren't used for providing library routines like zlib. The kernel does not link .so files. The code is almost certainly cut-and-pasted into the ppp compression code somewhere.
Wrong, too. The aspect ratio of the movie screen depends on the film format. 35mm is 16x9, but 70mm and 135mm are much wider. Most movies today are indeed shot in 35mm, but not all, particularly epic movies (Last of the Mohicans comes to mind).
"ratings 2-8 encompass 99% of films."
So in Plot Originality, Visual Impression, and Film Transfer, the DVD is better than 995 out of 1000 other DVDs (at least, I suppose his scale must be logarithmic or exponential?)
Sounds a little high.
I was one of the Dell laptop users whose batteries had a chance of catching on fire. Under Dell's replacement plan, they sent me a new battery immediately, and whenever I send back the defective one I get a second battery. That's a hell of a deal... I hardly use my batteries, it's mostly a desktop replacement, but I'm worried about the batteries dying completely someday and making the laptop plug-in only. I'm planning on saving the second battery until the first one dies. Thanks, Dell!
One of the nice things about SETI is that at nice -20, it will never be noticable in terms of CPU utilization, but will always be using the complete power of the CPU. Could we do that with disks?
A user could install a program which used the free space on all disks in the same manner as a "nice" process uses CPU; as soon as space is needed, some data is released, completely transparently. A company or organization could store data on the distributed network; they would keep a "master" copy of the data available, in case a particular fragment happened to be erased on all of the nodes, or nodes were unavailable.
The question I'm pondering is how to keep track of where data is stored, and route data from the nodes to the host where it would be read. In article's example, the fragments of a movie, sent to a particular client. How do we efficiently request fragments, in the correct order, without either overusing bandwidth with duplicated data or dropping fragments?
Fox can easily prevent X-Files watchers from acquiring copies of the episodes. Just don't broadcast them.
The good thing is that in courts, the argument of "if they do this it will hurt our business" doesn't hold up, even for baseball and it's strange exemption from antitrust laws.
Huh? Active Directory (and Exchange before that) is nothing but LDAP. DOM and HTML/HTTP are both obviously in IE. Kerberos is supported, certainly IMAP...
Microsoft has in fact been *catching up* to these standards for a few years now. Doesn't that sound interesting?
...it's a win for Intel. Larry says nothing in the article about the capabilities of Linux except that it's better than Windows "if you're on the Internet."
What he really liked, apparently, was the fact that the hardware was cheap and easily replaceable. It's a win for clustering, certainly, but is it a win for Linux?
My favorite story about Woz:
He was always a fan of interesting numbers, people who had addresses like 1234 Main Street. For a long time he wanted a phone number with all ten (or at least seven) digits the same, but couldn't get one. Finally when the 888 toll-free area code came out, he was able to get a cell phone with the number (888)888-8888.
Soon thereafter, he began getting mysterious calls on his new phone. The phone would ring and there would be just silence, or strange (but not particularly rude) noises. These happened several times a day. Eventually, in one of those calls, he heard a woman's voice: "What are you doing with that? Put that down." followed by the other end hanging up.
He figured out that it was babies who were calling him. If a baby or young child picks up the phone, one of the most likely numbers to dial is the same digit over and over. Kids were picking up the phone and mashing the 8 button constantly.
I read this in a Wired interview (doesn't seem to be online) which ended with the line, "...the babies of the world were calling the Woz."
You can't get your Jack Bot pregnant, so there's no need to order a new Rubber.
Well, no, it was Linus himself who first paired the kernel with GNU tools. He developed it using GNU development tools on Minix and then Linux itself once it was usable, and ASAP he had the standard GNU toolset compiled and running on his own box.
Perhaps Slackware did indeed release the first distribution; but from the start, GNU was used on Linux. I imagine that the 0.1 release had some mention of "BTW, if you want to actually DO anything, go get the GNU stuff..."
For one thing it's a quarter million, not a quarter billion, and for another I'm never going to be impressed by a number of lines of code, but by how well it works.
What do you mean, it wasn't literal? Clarke and Kubrick obviously thought about things they thought would be happening in the near future. I seem to recall Clarke being pessimistic about an AI as smart as HAL, but that's not quite enough to label the date of 2001 as "not literal." In the book, the events clearly happen in the year 2001 AD (or most of them, anyway). 2001 is much more specific and literal than a dystopian book like 1984 (where I would agree the date is more symbolic).
Science fiction is never completely accurate, obviously. But Clarke was one of the most accurate and scientifically rational writers of the century. We haven't gotten to convenient interplanetary travel quite yet, but you can be sure that it will happen much like he describes: a large space station using 'centrifigal force' to simulate gravity, and rockets using the station as a waypoint so the same spacecraft doesn't have to be capable of lifting off from Earth as well as travelling to and landing on another planet or moon.
Now, being able to phone from the station to America for only a few dollars, that's probably a little over-optimistic...
"Never underestimate the bandwidth of a station wagon filled with magtape, or a 747 filled with CD-ROMs"
I believe that filenames can even include /. Shells might not like it, but Unix filesystems don't care.
There's a bird called the Tufted Titmouse. The name for a group of these birds is a "rack".
Setun operated on numbers composed of 18 ternary digits, or trits
Awww...they shied away from the obvious choice, tits.
NT is certainly nothing like what it was originally intended to be. I've heard that it was developed on Alphas and then ported to the i386, just to make sure the developers made portability a priority. Look how long THAT lasted...
However, NT is POSIX compliant, just as much as Linux is, in fact. POSIX is a very general and practically useless standard; it's very easy to implement because it defines very little, and leaves many important considerations out.
All data is NOT WRITTEN TWICE. RAID is a solution for replicating data. Journalling file systems replicate meta-data, which is the information ABOUT a file, such as its name and where it's stored on disk. They eliminate the need for fsck, which will not recover lost data either. Before you bash it, understand it.
"Winux."
If I were writing an application which had a hope of running remotely (a standard windowed word processor for example) I would write it to X. But if I were writing a new flight simulator, I would know up front that there is no hope of running it remotely, because it needs direct hardware acceleration, and I would write it for the DirectFB layer.
This is more like DirectX than anything; a way to bypass the high-level windowing system to write directly to hardware. As people have said, it doesn't replace X completely. But I would rather have a X server layer on top of a direct-hardware layer, than a direct hardware piece hacked into an old X server.
The problem with DRM is the same as with computer security. It's very easy to secure a computer so that no one can use it. It's very difficult to secure a computer that a great many people need to use in a great many ways, while at the same time restricting any unauthorized users whatsoever.
Media companies really want to stop unauthorized use of their copyrighted material. Copying a CD and giving it to someone else is illegal. Copying a CD or creating a compilation for your own personal use is legitimate. The problem is allowing the legal copies while preventing the illegal ones. And it's a very difficult one to solve. It's much easier to simply prevent all copies whatsoever.
Anyone who thinks that "fair use" means giving away copies of music or books is a thief and an idiot. Remember this: if you stop paying for your media, they will stop selling it to you.