Western sensibilities tell us that there is a difference between a combatant and a civilian. Yet no one has told that to the other side.
To first answer your question:
YES we must maintain this distinction, as difficult as it may be at times, else we are savages and (importantly) are not doing good. You know the phrase "we had to destroy the village to save it", right?
During WW2 we didn't really care and perhaps that guilted some people.
It didn't just "guilt" some people: it also resulted in the slaughter of millions of innocents, for instance in the firebombing of Japanese and German cities (Tokyo, Kyoto, Dresden, etc.), and of course the atomic bomb.
That is, what is the moral difference between bombing a city and Hitler's Final Solution? That is, they are both military actions against civilian targets with marginal military benefits, and cause horrific suffering. (there clearly is a difference, but I'm suggesting that indiscriminate killing of civilians is not much better than genocide)
In Viet-nam we also didn't respect a civilian/combatant distinction, and what did it get us? 1 million Viet deaths and 50,000 American -- for what?
This points out another problem with indiscriminate killing of civilians: you fail to win the "hearts and minds" of the opponent, and cannot secure a lasting peace. Consider for instance in the US Civil War, Sherman's March to the Sea, when he laid waste to the South,
causing awful suffering, humiliating it, and undermining the possibility of Reconstruction.
The problem with separating the two is that in the long run the wars are prolonged and so is the suffering.
This is known as the "war is hell" doctrine: war is so bad that anything that shortens it is worth the cost. This is firstly a cop-out: it denies the need to apply moral decisions to war. Secondly, it is generally in retrospect seen to excuse activities (targetted killing of civilians, rape, pillage) that do not materially affect the success of war.
One of the great lies of the 20th century were systems (like Fascism and Communism) which said: "We must kill millions of innocents to reach the promised land". We have seen where that leads: a hecatomb of corpses and no promised land.
For a far better discussion of these issues (and you certainly seem interested), please read "Just and Unjust Wars", by Michael Walzer, where he discusses the issue of ethics in war in great detail and lucid prose. It's widely recognized as a masterpiece in the area.
Your subject line is rather expressive: I think that the goal (of the webpage author) is to communicate his understanding. More specifically, he wants to communicate it visually, but this is too narrow a goal. (An analogy, perhaps: someone building a housing project may say that their goal is to build a skyscraper, but their actual goal is (or should be) to house the poor. (yeah, I'm in Chicago)). So while the author succeeded in his narrow goal of drawing pictures to explain the Strouhal number, he failed (or at least did a suboptimal job) to convey the understanding.
There's something else here: it's often easiest to explain some idea by showing the way you discovered it, but this is usually bad for communicating and reflects a lack of understanding: if you understand something, you can often explain it more simply than by reciting the steps you used. So after drawing some pictures, the author realized that the Strouhal number was just the ratio of vertical to horizontal (speed, or distance/cycle) -- but instead of saying that, concisely, he draws the pictures and only then, awkwardly, explains the ratio.
That is, he put a lot of work into presenting his understanding, but some of that work should have been spent mulling over his understanding and polishing it, rather than presenting a gorgeously illustrated first draft.
Quoting now:
I wanted to see what ratios of amplitude, frequency, and speed actually look like in winged flight, what the Strouhal number actually represents, and why it is dimensionless.
As you note,
"look like" and "actually represents" indicate what kinda goal he has -- I would add "see" also suggests his visual goal. However, he's expressing it in visual terms b/c he's a visual person: I argue that his ultimate goal is understanding: "why it is dimensionless", say.
Incidentally, I am a geometer, hence my especial interest in this story: I'm also very visual. However, I think that visual understanding happens in your mind's eye, not in actual pictures. Perhaps this is because I do somewhat abstract geometry (I specialize in high dimensions, mainly 5 or more), hence we can't draw actual pictures, yet we can still use our visual intuition. Drawing pictures can help, but the actual understanding is not in the pictures: it's in your brain. Thus paradoxically pictures can inhibit visual understanding: you see the picture on the page (er, screen), but you don't have it in your head as an understanding.
Similarly, he essentially says that the Strouhal number is unitless because the units cancel, instead of saying that it's a ratio of speeds. Here we have a similar situation to pictures, only with equations: manipulating an equation can give you a result, just like you can see a result in a picture, but it doesn't mean you necessarily understand it, or can convey that understanding.
Note that I also teach Calculus, so exposition is often on my mind: I was chastising the author, not because he did a bad job, but because it could have been rather better.
Oh, and about Hilbert: his goal to eliminate pictures from mathematics was quixotic: turns out language is as squishy as pictures. If you want completely rigorous mathematics, you can do it (as Goedel and some did), but by making it machine-readable, it's no longer human readable. A quip about geometry is that it's "precise reasoning about imprecise pictures": the pictures are essentially schematics -- but so are the words: reality is elsewhere.
Note: I'm a mathematician (not a scientist, say), and I was remarking on the webpage, not the Nature article.
That is, the webpage has lots of pretty pictures to explain the Strouhal number, but doesn't provide a concise description: he essentially says as much (as another poster notes), but I clearly helped some other readers understand -- and by writing a paragraph, not drawing 200 pictures. This (that the webpage, while flashy, didn't communicate terribly well) was my implicit message; my explicit message was that the Strouhal number is easy to understand, and has a pretty obvious connection with "propulsive efficiency".
The exact connection between the Strouhal number and efficiency is commented on in the
abstract:
Propulsive efficiency is high over a narrow range of St and usually peaks within the interval 0.2 < St < 0.4 (refs 3-8). Because natural selection is likely to tune animals for high propulsive efficiency, we expect it to constrain the range of St that animals use.
This is what I meant by "not too surprising". Note however that the authors also write:
This [range] seems to be true for dolphins, sharks and bony fish, which swim at 0.2 < St < 0.4. Here we show that birds, bats and insects also converge on the same narrow range of St, but only when cruising.
In other words, this was know for swimmers, and here they've shown it for fliers, but only when cruising. As others have noted, birds have other types of flight that are not so tuned: soaring birds of prey, or hovering humming birds, for instance.
So: the Nature article is fine: saying that a result is "not surprising" is not a criticism in science (that's how you form a hypothesis). I do not presume to dispense with data, simply provide insight. Further, just because something is plausible doesn't make it true -- hence the need (and value) of experiments in science. OTOH, just because something is true doesn't mean we understand it -- hence the need for theory and intuition, which I was trying to provide.
The webpage, OTOH, while pretty, does a poor job of communicating. One paragraph and two pictures (implicit in my description) could have given insight into why Strouhal numbers fall in a narrow range.
The author of the webpage uses far too much verbiage and pretty pictures and doesn't seem to understand what's happening.
Amplitude of wing stroke times frequency of wing stroke is (half) the vertical speed of the wing tip. Forward speed is obviously the horizontal speed. Hence the ratio is dimensionless, and measures how steep (on average) the strokes are. It's not too surprising that these ratios should fall in a relatively narrow range, and the authors of the Nature article say as much.
Concretely, if the ratio were, say, 100, the wing would be flapping up and down furiously without advancing much -- if the ratio were 1/100, the wing would hardly be moving.
This question was made for me! My main box is a Pentium Pro 200 from 1996.
I have just finished upgrading my Dell Dimension XPS Pentium Pro200n to RedHat 9. Why did it take me so long? Well...my floppy drive was a bit wonky, so I couldn't use a boot floppy and I wanted to boot off CD. But ISOLINUX was having trouble booting off my CD, because the BIOS was so old that I couldn't run El Torito correctly. Turns out (googling revealed someone else from March of 2003 with the same computer and same problem) that a BIOS flash would work -- except of course my floppy drive didn't work, so I couldn't use the Dell-supplied program to make a boot disk (as my floppy drive didn't work -- though it took me a box of 10 new floppies before I decided that it was the drive and not just declining standards in floppies). Eventually I went in to school, made
a boot disk, then used 'dd' to extract the image, burned a bootable CD (see, bootable CDs that only used the 1.44 MB area booted okay, but ISOLINUX ones weren't), flashed the BIOS, installed RedHat 9, and I'm ready for another few years!
In retrospect I realize that I could probably have created a fake floppy drive under FreeDos (just create a filesystem-in-a-file and told it to mount as A:) to get the BIOS flash program without ever using a floppy drive. I now no longer have a floppy drive, or even a/mnt/floppy!
Also, I do plan to get a new computer within the year, but keep this old one (beatrice -- I name my boxen after characters from Dante) around as a file server and off-site backup.
The computer runs fine with GNOME, though no nautilus (mostly to save on memory). Mozilla/FireBird runs a bit slow, but that's my only complaint. It's really the Franken-system: only the motherboard, case, and power supply (and a few cards) are original. The hard drive has been replaced (twice), the CD-ROM became a CD-RW, I replaced the case fan, upgraded the memory, and I've even just replaced the monitor (old one has been getting a bit shaky/bouncy). However, 2MB of video memory doesn't go far on a new 19" monitor, so I've had to upgrade the video card too! (that and the S3 ViRGE (Number 9 Reality 332) has snow problems in XFree 4.3).
It's getting to sound like grandpa's hammer, where the head has been replaced 3 times and the handle twice, but it's still grandpa's hammer.
I know CD/DVDs can be read on any platfom, but not HDDs. Last time I tried that, I had to cut my 40GB disk in 2, since FAT32 doesn't support more than about 32GB partitions.
FAT32 supports partitions up to 2TB, but newer versions of Windows won't format them that large, though it can read them (this is to "encourage" people to use NTFS). Use Linux mkdosfs or Win98 or WinME, and you can format larger that 32GB.
Last week I bought a 120GB external hard drive, and formated it as one FAT32 partition, using mkdosfs (I've since resized it to have a "small" ReiserFS partition for backups: I want to preserve symbolic links and permissions!). I've just ordered another, and I plan to backup by using both drives, and storing one off-site. This eliminates most backup concerns.
Regarding read-only, under Linux you can mount filesystems read-only; this is presumably less reliable than a jumper setting, but should shield you from bad user-mode programs.
Joesph Louis Lagrange happened to be on the review council and refused to believe that adding sinusoids could produce signals with corners.
And they really can't; at a discontinuity the Fourier series converges to half the sum of the value of the function approached from the left and right...
You've confused discontinuity with corner, aka, non-differentiable point.
absolute value of x has a corner at 0, but is continuous there, and the fourier series for absolute value of x (on a suitable interval) converges at 0.
In fact:
Weierstrass first demonstrated an everywhere continuous but nowhere differentiable function as a sum of sinusoids.
On a closed and bounded interval (like [0,1], say), polynomials are dense in the space of continuous functions (in the uniform metric, if that helps), meaning that every continuous function (including those with lots of corners) is a limit of polynomials, which have no corners.
Here's another worst game:
Chiller, by Exidy, 1986.
It's a "horror shooter", which consists of you shooting people (mostly naked) in torture devices (and other stuff, but mostly torturing people).
you actually torture people. part of the game is turning a vice to crush someone's head; another is turning a rack to tear people apart. another is lowering someone into a river of blood so that
they can be torn apart by a croc. another is releasing a guillotine to cut someone's head off (and then shooting their face after their head's been cut off)
the people are very detailed: you often need to shoot the chest, abdomen, thigh, shin, foot, upper arm, forearm, hand and face all separately (points for each one), and if the face hasn't been shot yet, it reacts (groans or scream) to each shot.
The programmers are sick: they invested a lot of effort into programming how you can torture and mutilate people - and they're not "faceless drones" - they are a few stationary, helpless people.
It's the most vile and tasteless game I've ever seen. I've seen thousands, and this is by far the worst; I feel dirty for even looking at it.
You can check it out in MAME if you must;
here's a screenshot (scream-shot?) or two.
Diffie-Helman is not based on the knapsack problem (which is roughly: given a bunch of sack of various sizes (say, all of size 100) and a bunch of objects, what's the most efficient way of packing the objects into the (least number of) sacks?); DH is based on the discrete logarithm.
Note: people are interested in the knapsack problem because it is NP-complete to solve in general; the problem is that many (many!) specific cases are very easy to crack, and it's hard to tell a priori if you are generating such an example (a similar thing occurs in factor, in that there are some numbers that are much easier to factor than they may appear).
The discrete logarithm is as follow: recall that computing the logarithm (base b=10, say) of a number n is determining a number a such that 10^a=n. If you're working over the real numbers, this is easy to solve, and any calculator can do it quickly. On the other hand, suppose you are working in the integers modulo a prime p (imagine you're on a clock with p hours); then it's still possible to raise an integer b to a power a, getting a number n, and this is very quick. Say, given p=7 and b=5, and a=4, we get that 5^4 = 25*25 = 4*4 = 2 (modulo 7), so the discrete log base 5 mod 7 of 2 is 4. This is what Diffie-Helman relies on. Note that there are variants that work in different groups than this (a group is a mathematical object where you can add and subtract, roughly), particularly with elliptic curves, and these last are much touted as being possibly more secure than RSA or DH.
Where do I set the colour depth and resolution of my display
Where do I share things ?
Where do I reconfigure my network settings ?
These are all dealt with (and many more: how do you manage users?) in Ximian Setup Tools, a collection of programs (generally perl scripts) that provide an abstraction from the underlying system and allow one to configure the above via command-line or point-and-drool GUI tools. Incidentally, they are working with KDE on the backend to prevent duplication, so all that one needs to do is slap a pretty face on it. On the GNOME side, this is being integrated into the new control center (check out Red-Carpet Ximian Preview). Many of these tools need work/polish, but they're coming along very nicely. This is a known problem, and is being addressed. Note that the solution is much more integrated/elegant than Window's ``Where do I go to change that?'' (oh, it's easy: open an explorer window, then go to View->File types... and try to navigate that, er or was it in the control panel?)
Where do I load and unload kernel modules ?
? These are dynamically loaded and unloaded as needed or not on Linux, so this prolly doesn't get as much attention.
Where can I reconfigure my kernel, compile it, isntall it and reboot all by checking a few boxes and hitting a button ?
You can't do this at all on non-free systems, and vendors generally ship with a kernel that is fine for most purposes/people. That said, with the new Linux kernel configuration system (curtesy ESR), one could do this, but the sentiment seems to be that if you're recompiling a kernel, you probably should know enough about computers to not be afraid of `make menuconfig', so this has not received much attention, afaik.
Oh, and with some finesse, you don't need to reboot to load a new kernel -- this would be kinda nice to have a GUI tool for;-)
I have read science fiction novels in which there are technologies that nobody remembers the origins of..nobody knows who's responsible for it.
This sounds similar to the Jargon File -- an amorphism file that got propagated and mutated over time. Consider what has happened with it: b/c it was generally useful, it aquired a maintainer (ESR) and a main stream.
I expect this to be the general case, for instance if some source code is found/circulated, if it is useful it will aquire some sort of maintainer (e.g., in the interactive fiction archive, I found a useful little program which couldn't compile with gcc, and whose author had moved on (their email bounce). I made a lil' patch and put myself down as a contact, since it's useful enough.). Thus, while the concept of an unknown maker is interesting, I expect in reality it would be restricted to minor works of limited interest, if only b/c it is relatively easy to pick up maintainership of a found object in this sense.
Seriously though, we'd assume that minitruth would be able to do a little search-and-replacing, every time allegences changed, no? Hmm....job openings for perl hackers in 1984? Was perl even around then?
RSA encryption will be one of the best types around, at least mathematically speaking.
This is incorrect. RSA is strictly worse than Diffie-Helman. To elaborate, there are four problems we need to distinguish:
RSAP, the RSA Problem: given a large composite number, factor it.
DHP, the Diffie-Helman Problem: given A in Z/p (integers modulo some prime p) and n a positive integer, find B such that B^n=A. This is called the `discrete logarithm' b/c if we have A and B in the real numbers, this is just finding B=log_n A.
RSA, a cryptalgorithm based on the RSAP
DH (Diffie-Helman), a cryptalgoritm based on the DHP
To elaborate: a solution for the DHP would give a solution to the RSAP; thus RSA is on shakier ground than DH. Note that if cracking DH or RSA does not require solving DHP (resp. RSAP), then this is academic. We thus get the diagram:
DHP >= DH
>=
RSAP >= RSA
that is, if you can solve DHP, you can solve all of these, but if you can solve RSAP, you only get RSA cracked. Thus the security of DH (resp. RSA) is based on two assumptions:
Cracking DH/RSA requires solving the underlying mathematical problem (DHP/RSAP)
The underlying mathematical problem is hard
Unless you think there is something wrong with DH that allows cracking it without solving DHP, then DH is better than RSA. That said, the reason RSA is so prevalent is b/c it was patented after DH, and thus RSA inc. could make $$$ off it longer than if they used the technically superior DH; however, RSA has been ``good enough''; regardless of whether governments can crack it, $|<r1p+ <1dd1e$ sure can't. But that's all RSA is: good enough.
RSA problem is NP, but not known to be NP-hard
on
RSA Cracked - Not
·
· Score: 1
You may know more about RSA than I do, but isn't the RSA algorithm NP? Or, is just O(x^)?
Yes, they are NP, that is, given two integers P, Q that are claimed to be a factorization of a given number N, you can compute their product P*Q and compare it to N in polynomial time. Remember, NP means ``an answer can be verified in polynomial time''; it does NOT mean `not P' (where P means: a solution can be found in polynomial time). NP is a superset of P.
That said, neither the RSA problem (factor large integers) nor the Diffie-Helman problem (compute discrete log) are proven to be NP-hard (that is, if you can solve them, you can solve any NP problem). Some attempts have been made to make crypto algorithms that are based on NP-hard problems (back pack packing, traveling salesman), but no satisfactory implementation of these is currently known.
Next I tried the Gnome/GTK+ client. After realising the dependencies were spiralling out of control
I gave up. Nobody should have to update their entire (up to date Helix) gnome install just to get a
jabber client running. OK, I'm exhagerating a little. But it was enough to put me off, and certainly
not something we could force on our customers.
I have also had difficulties with dependencies,
as I use Redhat 7.0; since Gabber is written in
C++ and depends on libraries written in C++,
the move to gcc `2.96' hit it hard. However, you
can get a statically linked binary of the latest
Gabber (0.8.0), which has no crazy dependencies,
from the download page (or the above link).
Slashdot Mirror
Should there be a difference?
Western sensibilities tell us that there is a difference between a combatant and a civilian. Yet no one has told that to the other side.
To first answer your question: YES we must maintain this distinction, as difficult as it may be at times, else we are savages and (importantly) are not doing good. You know the phrase "we had to destroy the village to save it", right?
During WW2 we didn't really care and perhaps that guilted some people.
It didn't just "guilt" some people: it also resulted in the slaughter of millions of innocents, for instance in the firebombing of Japanese and German cities (Tokyo, Kyoto, Dresden, etc.), and of course the atomic bomb.
That is, what is the moral difference between bombing a city and Hitler's Final Solution? That is, they are both military actions against civilian targets with marginal military benefits, and cause horrific suffering. (there clearly is a difference, but I'm suggesting that indiscriminate killing of civilians is not much better than genocide)
In Viet-nam we also didn't respect a civilian/combatant distinction, and what did it get us? 1 million Viet deaths and 50,000 American -- for what?
This points out another problem with indiscriminate killing of civilians: you fail to win the "hearts and minds" of the opponent, and cannot secure a lasting peace. Consider for instance in the US Civil War, Sherman's March to the Sea, when he laid waste to the South, causing awful suffering, humiliating it, and undermining the possibility of Reconstruction.
The problem with separating the two is that in the long run the wars are prolonged and so is the suffering.
This is known as the "war is hell" doctrine: war is so bad that anything that shortens it is worth the cost. This is firstly a cop-out: it denies the need to apply moral decisions to war. Secondly, it is generally in retrospect seen to excuse activities (targetted killing of civilians, rape, pillage) that do not materially affect the success of war.
One of the great lies of the 20th century were systems (like Fascism and Communism) which said: "We must kill millions of innocents to reach the promised land". We have seen where that leads: a hecatomb of corpses and no promised land.
For a far better discussion of these issues (and you certainly seem interested), please read "Just and Unjust Wars", by Michael Walzer, where he discusses the issue of ethics in war in great detail and lucid prose. It's widely recognized as a masterpiece in the area.
Your subject line is rather expressive: I think that the goal (of the webpage author) is to communicate his understanding. More specifically, he wants to communicate it visually, but this is too narrow a goal. (An analogy, perhaps: someone building a housing project may say that their goal is to build a skyscraper, but their actual goal is (or should be) to house the poor. (yeah, I'm in Chicago)). So while the author succeeded in his narrow goal of drawing pictures to explain the Strouhal number, he failed (or at least did a suboptimal job) to convey the understanding.
There's something else here: it's often easiest to explain some idea by showing the way you discovered it, but this is usually bad for communicating and reflects a lack of understanding: if you understand something, you can often explain it more simply than by reciting the steps you used. So after drawing some pictures, the author realized that the Strouhal number was just the ratio of vertical to horizontal (speed, or distance/cycle) -- but instead of saying that, concisely, he draws the pictures and only then, awkwardly, explains the ratio.
That is, he put a lot of work into presenting his understanding, but some of that work should have been spent mulling over his understanding and polishing it, rather than presenting a gorgeously illustrated first draft.
Quoting now:
I wanted to see what ratios of amplitude, frequency, and speed actually look like in winged flight, what the Strouhal number actually represents, and why it is dimensionless.
As you note, "look like" and "actually represents" indicate what kinda goal he has -- I would add "see" also suggests his visual goal. However, he's expressing it in visual terms b/c he's a visual person: I argue that his ultimate goal is understanding: "why it is dimensionless", say.
Incidentally, I am a geometer, hence my especial interest in this story: I'm also very visual. However, I think that visual understanding happens in your mind's eye, not in actual pictures. Perhaps this is because I do somewhat abstract geometry (I specialize in high dimensions, mainly 5 or more), hence we can't draw actual pictures, yet we can still use our visual intuition. Drawing pictures can help, but the actual understanding is not in the pictures: it's in your brain. Thus paradoxically pictures can inhibit visual understanding: you see the picture on the page (er, screen), but you don't have it in your head as an understanding.
Similarly, he essentially says that the Strouhal number is unitless because the units cancel, instead of saying that it's a ratio of speeds. Here we have a similar situation to pictures, only with equations: manipulating an equation can give you a result, just like you can see a result in a picture, but it doesn't mean you necessarily understand it, or can convey that understanding.
Note that I also teach Calculus, so exposition is often on my mind: I was chastising the author, not because he did a bad job, but because it could have been rather better.
Oh, and about Hilbert: his goal to eliminate pictures from mathematics was quixotic: turns out language is as squishy as pictures. If you want completely rigorous mathematics, you can do it (as Goedel and some did), but by making it machine-readable, it's no longer human readable. A quip about geometry is that it's "precise reasoning about imprecise pictures": the pictures are essentially schematics -- but so are the words: reality is elsewhere.
Note: I'm a mathematician (not a scientist, say), and I was remarking on the webpage, not the Nature article.
That is, the webpage has lots of pretty pictures to explain the Strouhal number, but doesn't provide a concise description: he essentially says as much (as another poster notes), but I clearly helped some other readers understand -- and by writing a paragraph, not drawing 200 pictures. This (that the webpage, while flashy, didn't communicate terribly well) was my implicit message; my explicit message was that the Strouhal number is easy to understand, and has a pretty obvious connection with "propulsive efficiency".
The exact connection between the Strouhal number and efficiency is commented on in the abstract :
Propulsive efficiency is high over a narrow range of St and usually peaks within the interval 0.2 < St < 0.4 (refs 3-8). Because natural selection is likely to tune animals for high propulsive efficiency, we expect it to constrain the range of St that animals use.
This is what I meant by "not too surprising". Note however that the authors also write:
This [range] seems to be true for dolphins, sharks and bony fish, which swim at 0.2 < St < 0.4. Here we show that birds, bats and insects also converge on the same narrow range of St, but only when cruising.
In other words, this was know for swimmers, and here they've shown it for fliers, but only when cruising. As others have noted, birds have other types of flight that are not so tuned: soaring birds of prey, or hovering humming birds, for instance.
So: the Nature article is fine: saying that a result is "not surprising" is not a criticism in science (that's how you form a hypothesis). I do not presume to dispense with data, simply provide insight. Further, just because something is plausible doesn't make it true -- hence the need (and value) of experiments in science. OTOH, just because something is true doesn't mean we understand it -- hence the need for theory and intuition, which I was trying to provide.
The webpage, OTOH, while pretty, does a poor job of communicating. One paragraph and two pictures (implicit in my description) could have given insight into why Strouhal numbers fall in a narrow range.
The author of the webpage uses far too much verbiage and pretty pictures and doesn't seem to understand what's happening.
Amplitude of wing stroke times frequency of wing stroke is (half) the vertical speed of the wing tip. Forward speed is obviously the horizontal speed. Hence the ratio is dimensionless, and measures how steep (on average) the strokes are. It's not too surprising that these ratios should fall in a relatively narrow range, and the authors of the Nature article say as much. Concretely, if the ratio were, say, 100, the wing would be flapping up and down furiously without advancing much -- if the ratio were 1/100, the wing would hardly be moving.
Really? Perhaps you haven't seen their hit single "Developers! Developers! Developers!" You're missing out.
If you like the song, you'll love the video!
This question was made for me! My main box is a Pentium Pro 200 from 1996.
I have just finished upgrading my Dell Dimension XPS Pentium Pro200n to RedHat 9. Why did it take me so long? Well...my floppy drive was a bit wonky, so I couldn't use a boot floppy and I wanted to boot off CD. But ISOLINUX was having trouble booting off my CD, because the BIOS was so old that I couldn't run El Torito correctly. Turns out (googling revealed someone else from March of 2003 with the same computer and same problem) that a BIOS flash would work -- except of course my floppy drive didn't work, so I couldn't use the Dell-supplied program to make a boot disk (as my floppy drive didn't work -- though it took me a box of 10 new floppies before I decided that it was the drive and not just declining standards in floppies). Eventually I went in to school, made a boot disk, then used 'dd' to extract the image, burned a bootable CD (see, bootable CDs that only used the 1.44 MB area booted okay, but ISOLINUX ones weren't), flashed the BIOS, installed RedHat 9, and I'm ready for another few years!
In retrospect I realize that I could probably have created a fake floppy drive under FreeDos (just create a filesystem-in-a-file and told it to mount as A:) to get the BIOS flash program without ever using a floppy drive. I now no longer have a floppy drive, or even a /mnt/floppy!
Also, I do plan to get a new computer within the year, but keep this old one (beatrice -- I name my boxen after characters from Dante) around as a file server and off-site backup.
The computer runs fine with GNOME, though no nautilus (mostly to save on memory). Mozilla/FireBird runs a bit slow, but that's my only complaint. It's really the Franken-system: only the motherboard, case, and power supply (and a few cards) are original. The hard drive has been replaced (twice), the CD-ROM became a CD-RW, I replaced the case fan, upgraded the memory, and I've even just replaced the monitor (old one has been getting a bit shaky/bouncy). However, 2MB of video memory doesn't go far on a new 19" monitor, so I've had to upgrade the video card too! (that and the S3 ViRGE (Number 9 Reality 332) has snow problems in XFree 4.3).
It's getting to sound like grandpa's hammer, where the head has been replaced 3 times and the handle twice, but it's still grandpa's hammer.
I know CD/DVDs can be read on any platfom, but not HDDs. Last time I tried that, I had to cut my 40GB disk in 2, since FAT32 doesn't support more than about 32GB partitions.
FAT32 supports partitions up to 2TB, but newer versions of Windows won't format them that large, though it can read them (this is to "encourage" people to use NTFS). Use Linux mkdosfs or Win98 or WinME, and you can format larger that 32GB.
Last week I bought a 120GB external hard drive, and formated it as one FAT32 partition, using mkdosfs (I've since resized it to have a "small" ReiserFS partition for backups: I want to preserve symbolic links and permissions!). I've just ordered another, and I plan to backup by using both drives, and storing one off-site. This eliminates most backup concerns.
Regarding read-only, under Linux you can mount filesystems read-only; this is presumably less reliable than a jumper setting, but should shield you from bad user-mode programs.
Imagine a beowulf cluster of these . . .
Shouldn't this read:
Imagine a RAID array of these
The Mac is not already there with vector graphics. The sexy scalable icons are raster graphics which are rescaled, not vector graphics.
Yes, with Quartz/Display Postscript you have a vector drawing model, but you need APIs so it's actually usable.
AFAIK, GNOME was the first desktop to have a vector theme for it -- on this, Free Software is ahead of the Mac.
Joesph Louis Lagrange happened to be on the review council and refused to believe that adding sinusoids could produce signals with corners.
And they really can't; at a discontinuity the Fourier series converges to half the sum of the value of the function approached from the left and right...
You've confused discontinuity with corner, aka, non-differentiable point. absolute value of x has a corner at 0, but is continuous there, and the fourier series for absolute value of x (on a suitable interval) converges at 0.
In fact:
Here's another worst game: Chiller, by Exidy, 1986.
It's a "horror shooter", which consists of you shooting people (mostly naked) in torture devices (and other stuff, but mostly torturing people).
The programmers are sick: they invested a lot of effort into programming how you can torture and mutilate people - and they're not "faceless drones" - they are a few stationary, helpless people.
It's the most vile and tasteless game I've ever seen. I've seen thousands, and this is by far the worst; I feel dirty for even looking at it.
You can check it out in MAME if you must; here's a screenshot (scream-shot?) or two.
Diffie-Helman is not based on the knapsack problem (which is roughly: given a bunch of sack of various sizes (say, all of size 100) and a bunch of objects, what's the most efficient way of packing the objects into the (least number of) sacks?); DH is based on the discrete logarithm.
Note: people are interested in the knapsack problem because it is NP-complete to solve in general; the problem is that many (many!) specific cases are very easy to crack, and it's hard to tell a priori if you are generating such an example (a similar thing occurs in factor, in that there are some numbers that are much easier to factor than they may appear).
The discrete logarithm is as follow: recall that computing the logarithm (base b=10, say) of a number n is determining a number a such that 10^a=n. If you're working over the real numbers, this is easy to solve, and any calculator can do it quickly. On the other hand, suppose you are working in the integers modulo a prime p (imagine you're on a clock with p hours); then it's still possible to raise an integer b to a power a, getting a number n, and this is very quick. Say, given p=7 and b=5, and a=4, we get that 5^4 = 25*25 = 4*4 = 2 (modulo 7), so the discrete log base 5 mod 7 of 2 is 4. This is what Diffie-Helman relies on. Note that there are variants that work in different groups than this (a group is a mathematical object where you can add and subtract, roughly), particularly with elliptic curves, and these last are much touted as being possibly more secure than RSA or DH.
These are all dealt with (and many more: how do you manage users?) in Ximian Setup Tools, a collection of programs (generally perl scripts) that provide an abstraction from the underlying system and allow one to configure the above via command-line or point-and-drool GUI tools. Incidentally, they are working with KDE on the backend to prevent duplication, so all that one needs to do is slap a pretty face on it. On the GNOME side, this is being integrated into the new control center (check out Red-Carpet Ximian Preview). Many of these tools need work/polish, but they're coming along very nicely. This is a known problem, and is being addressed. Note that the solution is much more integrated/elegant than Window's ``Where do I go to change that?'' (oh, it's easy: open an explorer window, then go to View->File types... and try to navigate that, er or was it in the control panel?)
? These are dynamically loaded and unloaded as needed or not on Linux, so this prolly doesn't get as much attention.
You can't do this at all on non-free systems, and vendors generally ship with a kernel that is fine for most purposes/people. That said, with the new Linux kernel configuration system (curtesy ESR), one could do this, but the sentiment seems to be that if you're recompiling a kernel, you probably should know enough about computers to not be afraid of `make menuconfig', so this has not received much attention, afaik.
Oh, and with some finesse, you don't need to reboot to load a new kernel -- this would be kinda nice to have a GUI tool for ;-)
I have read science fiction novels in which there are technologies that nobody remembers the origins of..nobody knows who's responsible for it.
This sounds similar to the Jargon File -- an amorphism file that got propagated and mutated over time. Consider what has happened with it: b/c it was generally useful, it aquired a maintainer (ESR) and a main stream.
I expect this to be the general case, for instance if some source code is found/circulated, if it is useful it will aquire some sort of maintainer (e.g., in the interactive fiction archive, I found a useful little program which couldn't compile with gcc, and whose author had moved on (their email bounce). I made a lil' patch and put myself down as a contact, since it's useful enough.). Thus, while the concept of an unknown maker is interesting, I expect in reality it would be restricted to minor works of limited interest, if only b/c it is relatively easy to pick up maintainership of a found object in this sense.
Now cathedrals, on the other hand...
Now go listen to what Linux Torvalds is saying
Er, Linux Torvalds?
ROTFL!
Seriously though, we'd assume that minitruth would be able to do a little search-and-replacing, every time allegences changed, no? Hmm....job openings for perl hackers in 1984? Was perl even around then?
What? Not:
?</ob AYBABTU>RSA encryption will be one of the best types around, at least mathematically speaking.
This is incorrect. RSA is strictly worse than Diffie-Helman. To elaborate, there are four problems we need to distinguish:
To elaborate: a solution for the DHP would give a solution to the RSAP; thus RSA is on shakier ground than DH. Note that if cracking DH or RSA does not require solving DHP (resp. RSAP), then this is academic. We thus get the diagram:
that is, if you can solve DHP, you can solve all of these, but if you can solve RSAP, you only get RSA cracked. Thus the security of DH (resp. RSA) is based on two assumptions:
Unless you think there is something wrong with DH that allows cracking it without solving DHP, then DH is better than RSA. That said, the reason RSA is so prevalent is b/c it was patented after DH, and thus RSA inc. could make $$$ off it longer than if they used the technically superior DH; however, RSA has been ``good enough''; regardless of whether governments can crack it, $|<r1p+ <1dd1e$ sure can't. But that's all RSA is: good enough.
You may know more about RSA than I do, but isn't the RSA algorithm NP? Or, is just O(x^)?
Yes, they are NP, that is, given two integers P, Q that are claimed to be a factorization of a given number N, you can compute their product P*Q and compare it to N in polynomial time. Remember, NP means ``an answer can be verified in polynomial time''; it does NOT mean `not P' (where P means: a solution can be found in polynomial time). NP is a superset of P.
That said, neither the RSA problem (factor large integers) nor the Diffie-Helman problem (compute discrete log) are proven to be NP-hard (that is, if you can solve them, you can solve any NP problem). Some attempts have been made to make crypto algorithms that are based on NP-hard problems (back pack packing, traveling salesman), but no satisfactory implementation of these is currently known.
AFAIK, the change that justified moving the linux version number to 2.0 was that it was porting to multiple architectures (i.e., not just i386).
Next I tried the Gnome/GTK+ client. After realising the dependencies were spiralling out of control I gave up. Nobody should have to update their entire (up to date Helix) gnome install just to get a jabber client running. OK, I'm exhagerating a little. But it was enough to put me off, and certainly not something we could force on our customers.
I have also had difficulties with dependencies, as I use Redhat 7.0; since Gabber is written in C++ and depends on libraries written in C++, the move to gcc `2.96' hit it hard. However, you can get a statically linked binary of the latest Gabber (0.8.0), which has no crazy dependencies, from the download page (or the above link).
Why is it assumed that commercial software and free software must exist in opposition to one another? Simply because this is RMS's position?
There is no such dichotomy, as is made quite clear in the GNU definition of commercial software, nor does RMS adhere to such a belief.
This is easy to see today, since there are a large number of companies writing commercial free software (IBM, Red Hat, Helixcode, etc.).
To refer to non-free software, the FSF suggests using ``propriety'', though I prefer the term slave-ware.