Video - MPlayer, it even runs without X Window. Can Windows Media Player run video in MS-DOS?
What kind of argument is this? Who cares about DOS? WMP is OK when you install the missing codecs.
I can think of a few reasons why this is makes MPlayer better:
I think a GUI is useless for a movie player. Movies are "meant" to be watched fullscreen, and the keyboard is just fine for controlling them; it is just like the controls on a DVD player. I installed MPlayer on a friend's Windows machine and she doesn't find the lack of GUI a problem at all. Movie files are associated with mplayer.exe so there's no need to run it from the command line.
As movies are watched fullscreen, who needs a windowing system? The lack of a WS could mean the difference between playing and not playing on a low-end (embedded?) system.
As of the codecs, their quality was just the reason for the above installation. But furthermore it means more freedom to watch even the kind of material that Microsoft doesn't want you to watch (I'm thinking of digital restrictions management).
Bandwidth means the range of available frequencies. For example, human hearing is from 20 Hz to 20 kHz, so the bandwidth is 19980 Hz. This also defines the range of frequencies required, for example, to transmit monophonic radio signals (In fact, the bandwidth required by radio is usually more than the hearing bandwidth, but simplistically speaking it's the same).
In any communication channel, the data rate (as measured in bits per second) is proportional to bandwidth, and it kind of makes sense to confuse the two (it was originally a hacker joke). The problem is that they are really separate things. For example, plain old telephone has a bandwidth of a few kHz, but the data rate can be more than a few kbps if there is good signal to noise ratio. To be precise:
Data rate in bps = log2 (1+S/N) * Bandwidth in Hz
(Shannon's Law)
..the only way concurrent programming is going to play a major role for the majority of software, I believe, is at the compiler and OS levels: The OS and compiler designers are going to have to do their utmost to transform single-threaded software to perform optimally in a multi-CPU environment-
This is already being done with decent compilers and decent languages. It requires that the language is suitably high-level (unlike C, for example) so that the compiler has more freedom to make optimizations. HLLs should be nicer for programmers in many other ways too.
The relevant experience I have is with Fortran 90 and the PGF compiler back in 2001, on a dual P3 machine. The compiler produced parallel code to utilize both CPUs and their vector instructions (such as SSE).
So, this has been done for ages already, but people need to ditch low-level languages like C for this scenario to work.
There is a company in California,Corridor Systems, that is developing systems at 2.4 and 5.3GHz for BPL. It has 250MHz of bandwidth, and Shows VERY little interference.
What? Somebody on/. is using the word 'bandwidth' in its original and correct meaning? Unpossible!
It's the end of the world I say! CAT6Es and dogs living together!
And imagine using Dasher for coding - Dasher works well for writing words, but fails totally with the symbols and syntax used in programming.
Not necessarily. Dasher needs 'training' with the kind of text you want to write. Then it can use the probabilities with which a certain letter comes after another. These can be any symbols, they're not limited to letters of the alphabet. You could use different training texts for different kinds of task.
In fact, my (limited) experience with Dasher reminds me of the Windows start menu; it could be dasherized to make the most likely application bigger. And you could probably dasherize many other sequential tasks besides typing.
On the other hand, my fundamental problem with Dasher is that it promotes an ordinary, routine way of things, by emphasizing the most likely route you've used in the past. I would never use it for any kind of creative writing.
Agreed. A few months ago I was looking for something big and cheap and got this for 473 euros.
I do most of my real work on an old 800x600 laptop, and I find that a bigger screen only tends to be cluttered with more windows. I prefer a smaller screen and virtual desktops, and I wanted a bigger screen mainly for watching videos.
I think a better resolution would only be beneficial if it was something like 200 dpi or better, at which point you didn't need AA any more.
I think scientists have talked about paradigms a long time before 1994, and for good reasons. I can understand its overuse in business presentations though, as I'm annoyed by buzzwords in general.
I second the idea that multicore chips require higher-level languages than something like C. I've written some number-crunching stuff in Fortran 90, and its built-in matrix operations are incomparable; the compiler takes care of parallelizing for multiple CPUs and possibly vector instructions (like SSE).
When doing matrix operations in C, you have to write is as a loop. There are parallelizing compilers for C, but they have to take the low-level C loop back into a higher-level representation in order to parallelize. So there are more steps, some of them backwards, which is kind of dumb.
AFAIK hibernate feature on linux is not very stable yet.. (or I might be wrong)
It depends, of course. Some laptops have BIOS functions for suspend (to RAM) and/or hibernation (to disk), and the Linux kernel feature is yet another thing. On my old Toshiba Satellite, APM BIOS suspend (i.e. "apm -s") works fine, while I haven't got the kernel thingie to work on any of my machines.
Actually, DHCP is especially useful on a small network, as you can setup the DHCP server to assign addresses to specific MAC addresses instead of manually configuring the details in the OS.
At the school where I work, we do exactly this. The idea is that workstations can be exact copies of a master image, but of course each must have a unique IP address. It's a lot nicer to set these up at the server, than going through the boxes separately. Especially when some of them are Windows machines.
It's Hydrogen Hydroxide. H2O is wrong, it HOH, hence it having polar bonds. If you are going to make snarky comments, at least get them right.
First of all, as another poster already pointed out, dihydrogen monoxide is the name most often used in parodies. Many compounds are known by more than one name.
Moreover, your reasoning is wrong IMHO, because HOH could represent a non-polar molecule if it were symmetric like this:
H - O - H
But as we both know, it has a V shape, with the oxygen at the bottom of the V (haven't drawn it, as the damn ecode doesn't work with spaces like pre).
In fact, carbon dioxide has the structure O - C - O, but it's not polar because it's laid out in a straight line instead of a V.
I think H2O represents the structure of water better than HOH, because the two hydrogens are clearly on one side of the oxygen, not symmetrically apart. On the other hand HOH does represent the order of bonding better, but we're really talking about molecular, not structural formulae here.
Ogg is more efficient than mp3 on lower bitrates (without joint tricks)
Data compression is all about finding redundancy and encoding it more efficiently. In most music you find instruments/voices that are evenly spread over both channels, or with a fixed panning ratio across them. This is redundancy, and any sensible compression algorithm (lossy or not) will try and tackle it. With music compression this happens to be called joint stereo.
The depth edge maps bear a superficial resemblance to phase congruency maps. It's the best edge detection method I've come across, and works on ordinary 2D images. Check out some examples on Peter Kovesi's pages, there's also some code for download.
Slashdot Mirror
I could see them doing something like "commercialized internet" like they did for GPS.
Email is pretty much the same age as Internet, which is closer to 35 than 25 years.
On the other hand, innovation means roughly the same thing as commercialization, I think...
I can think of a few reasons why this is makes MPlayer better:
In any communication channel, the data rate (as measured in bits per second) is proportional to bandwidth, and it kind of makes sense to confuse the two (it was originally a hacker joke). The problem is that they are really separate things. For example, plain old telephone has a bandwidth of a few kHz, but the data rate can be more than a few kbps if there is good signal to noise ratio. To be precise:
See also: Throughput vs Bandwidth and links on that page.This is already being done with decent compilers and decent languages. It requires that the language is suitably high-level (unlike C, for example) so that the compiler has more freedom to make optimizations. HLLs should be nicer for programmers in many other ways too.
The relevant experience I have is with Fortran 90 and the PGF compiler back in 2001, on a dual P3 machine. The compiler produced parallel code to utilize both CPUs and their vector instructions (such as SSE).
So, this has been done for ages already, but people need to ditch low-level languages like C for this scenario to work.
What? Somebody on /. is using the word 'bandwidth' in its original and correct meaning? Unpossible!
It's the end of the world I say! CAT6Es and dogs living together!
Not necessarily. Dasher needs 'training' with the kind of text you want to write. Then it can use the probabilities with which a certain letter comes after another. These can be any symbols, they're not limited to letters of the alphabet. You could use different training texts for different kinds of task.
In fact, my (limited) experience with Dasher reminds me of the Windows start menu; it could be dasherized to make the most likely application bigger. And you could probably dasherize many other sequential tasks besides typing.
On the other hand, my fundamental problem with Dasher is that it promotes an ordinary, routine way of things, by emphasizing the most likely route you've used in the past. I would never use it for any kind of creative writing.
..is usually less when using a laptop, thus it benefits more from finer resolutions.
I do most of my real work on an old 800x600 laptop, and I find that a bigger screen only tends to be cluttered with more windows. I prefer a smaller screen and virtual desktops, and I wanted a bigger screen mainly for watching videos.
I think a better resolution would only be beneficial if it was something like 200 dpi or better, at which point you didn't need AA any more.
I think scientists have talked about paradigms a long time before 1994, and for good reasons. I can understand its overuse in business presentations though, as I'm annoyed by buzzwords in general.
Try not to anthropomorphize computers. They hate it when you do that.
Surely they must be safe, they're digitally signed and everything...
"Ani" is also the plural of "anus". As if "Exploiting the ANI hole" didn't sound funny enough already ;)
And I ain't gots the time to learn to spell.
Now with extra alkaloids!
http://torrentsearch.com/
When doing matrix operations in C, you have to write is as a loop. There are parallelizing compilers for C, but they have to take the low-level C loop back into a higher-level representation in order to parallelize. So there are more steps, some of them backwards, which is kind of dumb.
Of course, your favourite binary distro releases it without spending any compilation time... or any testing/management time for that matter...
AFAIK hibernate feature on linux is not very stable yet.. (or I might be wrong)
It depends, of course. Some laptops have BIOS functions for suspend (to RAM) and/or hibernation (to disk), and the Linux kernel feature is yet another thing. On my old Toshiba Satellite, APM BIOS suspend (i.e. "apm -s") works fine, while I haven't got the kernel thingie to work on any of my machines.
NMI
linky
At the school where I work, we do exactly this. The idea is that workstations can be exact copies of a master image, but of course each must have a unique IP address. It's a lot nicer to set these up at the server, than going through the boxes separately. Especially when some of them are Windows machines.
First of all, as another poster already pointed out, dihydrogen monoxide is the name most often used in parodies. Many compounds are known by more than one name.
Moreover, your reasoning is wrong IMHO, because HOH could represent a non-polar molecule if it were symmetric like this:
But as we both know, it has a V shape, with the oxygen at the bottom of the V (haven't drawn it, as the damn ecode doesn't work with spaces like pre).In fact, carbon dioxide has the structure O - C - O, but it's not polar because it's laid out in a straight line instead of a V.
I think H2O represents the structure of water better than HOH, because the two hydrogens are clearly on one side of the oxygen, not symmetrically apart. On the other hand HOH does represent the order of bonding better, but we're really talking about molecular, not structural formulae here.
Data compression is all about finding redundancy and encoding it more efficiently. In most music you find instruments/voices that are evenly spread over both channels, or with a fixed panning ratio across them. This is redundancy, and any sensible compression algorithm (lossy or not) will try and tackle it. With music compression this happens to be called joint stereo.
The depth edge maps bear a superficial resemblance to phase congruency maps. It's the best edge detection method I've come across, and works on ordinary 2D images. Check out some examples on Peter Kovesi's pages, there's also some code for download.