If I press the "mode" button on my c. 1991 Sony digital watch, it switches modes (Time/Stopwatch/etc.), but if I hold it down for three seconds, it launches the time-setting application.
A digital watch has a CPU, register, memory, and an OS -- it's the ultimate limited power computing device.
Good for you. The most important thing about perl is to remember to use its power for good, never evil. By that, I mean that perl merely gives you a mindblowingly powerful toolbox; it's your responsibility to write clear, maintainable code with those tools. Clarity is, unfortunately, much harder to achieve than hacked-together utility.
I remember the foofooraw a few years ago when some mathematician pointed out the blindingly obvious: not only are there far more irrational numbers than rational ones, but most of them are far too complicated ever to be named by mathematicians. After all, written mathematical expressions consist of sequences of symbols, which are isomorphic to the integers (a fact that computer scientists have used to great advantage by representing type as huge integers) -- but there are only countably many integers, so you can name at most a countable subset of the irrational numbers.
Yep. That changes the weighting between songs and randomness -- after a long time, it plays mainly things that you request a lot, whenever you're logged in.
I've been thinking about maintaining two separate lists, for more control over the random-vs-requested ratio, but the current naive scheme seems to work OK.
Our jukebox server keeps a list of all the requests each user makes. When the request list is empty, it randomly selects from a pool that includes the name of every song on the archive, plus the request lists. That way you get a mix that includes some truly random stuff but is weighted toward your favorites.
It produces a stream with the same appeal as a college radio station -- loosely aligned with a particular format, but quirky and eclectic.
I was about to post something about that until I saw your insightful post. No mod points, so, er, if you have 'em --
Re:Antivirus SW redundant for open-source...
on
GNOME for Grandma
·
· Score: 1
That's where there's money to be made! An ideal open-source "antivirus" company would release bugfix-patches to supported software, rather than prefilters that prevent exploits by intercepting them.
Think of it as a roof-patching company to stop up leaks in roofs. Norton et al. just build a second roof over the top of the original one, rather than patching up the holes in the original roof.
Capitalism works in large part because it rewards those who can create wealth, with a larger amount of control over society -- hence all of society benefits. The benefits of capitalism come from people wanting to get wealthy, and creating wealth out of good ideas and unexploited resources. The ills of capitalism come from the second way of getting rich: stealing or appropriating others' wealth, without creating anything of value.
The GPL is specifically designed to avoid (2), while allowing (1), in the context of pure information.
Antivirus SW redundant for open-source...
on
GNOME for Grandma
·
· Score: 1
... here's why. Antivirus software is a patch that protects vulnerable proprietary software from exploits. Antivirus companies make their living from the long delay between problems being identified in commercial closed-source software and those problems being fixed by the vendor. Open-source software can be patched locally, in the existing source -- hence a company that would make anti-exploit software for, say, LookOut would instead make patches for Ximian. Whether that is a business model that works remains to be seen.
The problem here is that the ``password'' (the port knock sequence) is sent in plaintext. Anyone with a sniffer anywhere between you and the other machine can see what you're doing. If this ever catches on, any L337 |1dd13 with half a rootkit will be sniffing for anomalous port-requests, and you'll be just as hosed as if you logged on via telnetd.
What a crappy marketing move -- naming a new material like another well-known material that is almost entirely unlike it. Holy fuck, people will be saying things like "Hey, could you go get the sapphire detector?" and they won't know if it's an ultraviolet-sensitive CCD or a leak-sensor for the damned fire suppression system.
Imagine the fun they'll have with it in Austin Powers 3. "No, you idiot, I wanted the sapphire(tm), not the sapphire."
I have that kind of misreporting. The implication in both the article and the press release is that wireless networks (such as cell phone networks) will be more secure because of these untrusting CPUs. But anyone seeking to hack the cell phone network won't work at the cell phone level -- they'll work at the transport portion of the protocol, sending bogus packets straight to base stations. The "secure CPU" won't have anything to do with it.
One may also entertain serious doubts about the airtightness of the CPU lockout. Other DRM platforms, such as Xbox, haven't exactly stood the test of time.
20th Street in Boulder, Colorado, has had such a speed-controlled light for years. Never thought it was a big deal until the self-important Caltrans guys announced theirs.
It works pretty well at regulating traffic, though not at boosting revenue for the city. Rather than build more, lately they've been parking a photo-radar van around town: that has the added benefit of filling the city's coffers.
ITAR applies to physical objects. EAR applies to aid and information. EAR does not apply to anything that is published (vs. proprietary). If you are developing and publishing a protocol suite or some open-source software, you are probably in the clear. The most obvious precedent is PGP (MIT Press went so far as to publish the source code as a book).
Did you notice that most of those missions were already launched? The budget projections for NASA are out for the next few years, and (at least for unmanned exploration and space science) they're not pretty.
NASA just cancelled an entire line of six spacecraft -- the Solar-Terrestrial Probes -- that have been on the drawing board since the mid 1990s. The Explorer line of missions is delayed indefinitely. Science funding is level for the next two years, then drops rapidly.
Meanwhile, countries like Japan, India, and China are building their space programs with vigor and dedication. Japan -- a nation the size of California -- will nearly match our rate of new scientific launches over the next decade.
The reason for the cuts in scientific launches at NASA is W's new manned-but-not-funded spaceflight initiative, which is diverting resources from the comparatively inexpensive scientific missions.
Actually, they will have a choice. My shop occasionally requires me to use M$O instead of Lyx/LaTeX, so I recently installed CodeWeavers' CrossoverOffice version of Wine, and installed Microsoft Office on top of that. It works perfectly. (Well, OK, pretty damned poorly -- but no worse than a native Windoze copy).
For cryin' out loud, I doubt that the music store will have much affect, as suggested by the poster -- after all, it's not a person and therefore doesn't have moods.
It will almost certainly have an "effect" -- that is, it will affect the market.
Am I the only slashdot reader who rankles at our editors' lack of good English? Ferchrissakes, folks, read Strunk and White -- it's online these days.
Hey, whaddaya know -- you're right! Sorry to say, though, the Wiki is at least partially wrong on this point. The RTGs on (for example) Cassini get most of their energy from alpha decay of the Pu-238 on board, but a significant number of both spontaneous and induced fissions also occur. (Wiki seems to say that no fission is occurring, perhaps in a bid to soothe activist nerves). Also, a significant amount of the energy comes from subcritical multiplication -- a non-self-sustaining chain reaction. The neutron-absorption -> fission cross section for Pu-238 is pathetic (less than 10% the cross section for Pu-239, which is what goes into nuclear weapons). But it's nonzero, so a measurable amount of chain reaction is occuring.
My mistake was in remembering subcritical multiplication being the dominant process when in fact it seems not to be for the RTG's we all know and love (the ones on Cassini).
You can check the cross sections and decay rates at the online chart of the nuclides, hosted in South Korea. (Until recently, that was hosted at BNL; does anyone else find it ironic that the Koreans are now exporting nuclear information to Berkeley?
Yup, chemical rockets can't get much past about 500 seconds because there just isn't any more energy in chemical bonds. The game in rocketry is to dump as much momentum as possible into as little propellant as possible. The rub: putting momentum into mass requires energy. Momentum scales linearly with your rocket's exhaust speed; but the kinetic energy of the exhaust scales like the square of the exhaust speed! Hence, the more propellant efficient you are, the more energy you need per unit mass. If I recall right, the most energetic-per-unit-mass reaction is atomic hydrogen bonding with atomic fluorine, yielding
0.1 eV/amu of fuel -- that translates to 9.6 megajoules per kg of fuel, or (with perfect conversion) about 4500 meters/second exhaust speed. Divide by 10 m/sec^2 to get Isp, and you find that 450 seconds is the limit for chemical rockets.
Nuclear reactions yield about a million times more energy per unit mass than do chemical reactions, so it's natural to try to get the energy that way.
NERVA got OK Isp (about a factor of 2 better than chemical rockets, something like 1000 seconds), but its thrust-to-weight ratio was pretty low, about 4 if I remember right. That's because it included a critical, operating nuclear reactor with an actively controlled chain reaction, and them thar things are heavy.
Thrust-to-weight is just as important as Isp to a rocket: higher thrust-to-weight means you can tote more fuel, payload, and structure for the same Isp, since you always have to have the mass of the engine itself around. By contrast to the NERVA's thrust-to-weight of about 4, the Space Shuttle main engines have a thrust-to-weight ratio of around 75. Since solid rockets are technically made out of their own fuel, their effective weight is much lower for this calculation (pretty much just the bell nozzle) and you might see numbers in the several-hundreds range.
Of course, one could always work on making the NERVA more lightweight -- but do you really want to optimize a nuclear reactor for mass, rather than safety? I didn't think so.
Now, for use in space, thrust-to-weight isn't so important. The rocket doesn't have to support itself against gravity, so low-mass engines that also produce low thrust are perfectly OK.
Of course, international treaty bans the use of critical nuclear reactors in space, but that alone wouldn't slow down our current administration very much.
[Nuclear reactors get flown into space all the time, but they always have much less than critical mass, relying on spontaneous decay to keep the chain reaction limping along at a constant rate. NERVA would require controlled reaction rates, hence a critical-mass reactor.]
Remember (nobody ever seems to) that getting into space is not nearly has hard as getting into orbit. We're used to descriptions of the amazing technology that is used to protect astronauts and such on re-entry -- but that amazing technology is only needed because of the enormous speeds that orbital vehicles have to attain.
The kinetic energy required to accelerate a gallon of gasoline to orbital speed is more than the chemical energy contained in the gasoline.
By contrast, "merely" lifting something up 100km doesn't require much energy at all.
So, er, no, leading-edge heat shields ought to be just fine. Fiberglass or carbon-fiber composites might even survive a flight or two without any shielding at all.
Slashdot Mirror
if I hold it down for three seconds, it launches the time-setting application.
A digital watch has a CPU, register, memory, and an OS -- it's the ultimate limited power computing device.
Good for you. The most important thing about
perl is to remember to use its power for good, never evil. By that, I mean that perl merely gives you a mindblowingly powerful toolbox; it's your responsibility to write clear, maintainable code with those tools. Clarity is, unfortunately, much harder to achieve than hacked-together utility.
Deep, heady stuff -- but who cares?
I've been thinking about maintaining two separate lists, for more control over the random-vs-requested ratio, but the current naive scheme seems to work OK.
It produces a stream with the same appeal as a college radio station -- loosely aligned with a particular format, but quirky and eclectic.
I was about to post something about that until I saw your insightful post. No mod points, so, er, if you have 'em --
Think of it as a roof-patching company to stop up leaks in roofs. Norton et al. just build a second roof over the top of the original one, rather than patching up the holes in the original roof.
The GPL is specifically designed to avoid (2), while allowing (1), in the context of pure information.
... here's why. Antivirus software is a patch that protects vulnerable proprietary software from exploits. Antivirus companies make their living from the long delay between problems being identified in commercial closed-source software and those problems being fixed by the vendor. Open-source software can be patched locally, in the existing source -- hence a company that would make anti-exploit software for, say, LookOut would instead make patches for Ximian. Whether that is a business model that works remains to be seen.
dratted HTML filter. That's ``L337 |<1dd13'', of course.
The problem here is that the ``password'' (the port knock sequence) is sent in plaintext. Anyone with a sniffer anywhere between you and the other machine can see what you're doing. If this ever catches on, any L337 |1dd13 with half a rootkit will be sniffing for anomalous port-requests, and you'll be just as hosed as if you logged on via telnetd.
Imagine the fun they'll have with it in Austin Powers 3. "No, you idiot, I wanted the sapphire(tm), not the sapphire."
One may also entertain serious doubts about the airtightness of the CPU lockout. Other DRM platforms, such as Xbox, haven't exactly stood the test of time.
It works pretty well at regulating traffic, though not at boosting revenue for the city. Rather than build more, lately they've been parking a photo-radar van around town: that has the added benefit of filling the city's coffers.
I am not a lawyer.
NASA just cancelled an entire line of six spacecraft -- the Solar-Terrestrial Probes -- that have been on the drawing board since the mid 1990s. The Explorer line of missions is delayed indefinitely. Science funding is level for the next two years, then drops rapidly.
Meanwhile, countries like Japan, India, and China are building their space programs with vigor and dedication. Japan -- a nation the size of California -- will nearly match our rate of new scientific launches over the next decade.
The reason for the cuts in scientific launches at NASA is W's new manned-but-not-funded spaceflight initiative, which is diverting resources from the comparatively inexpensive scientific missions.
Hamburg and Dresden were contributing to the war effort for Germany. Firebombing them sucked, but the "war crime" story started as German propaganda.
Actually, they will have a choice. My shop occasionally requires me to use M$O instead of Lyx/LaTeX, so I recently installed CodeWeavers' CrossoverOffice version of Wine, and installed Microsoft Office on top of that. It works perfectly. (Well, OK, pretty damned poorly -- but no worse than a native Windoze copy).
It will almost certainly have an "effect" -- that is, it will affect the market.
Am I the only slashdot reader who rankles at our editors' lack of good English? Ferchrissakes, folks, read Strunk and White -- it's online these days.
Switch to pine.
Or emacs/VM.
Or mutt.
Or...
My mistake was in remembering subcritical multiplication being the dominant process when in fact it seems not to be for the RTG's we all know and love (the ones on Cassini).
You can check the cross sections and decay rates at the online chart of the nuclides, hosted in South Korea. (Until recently, that was hosted at BNL; does anyone else find it ironic that the Koreans are now exporting nuclear information to Berkeley?
Nuclear reactions yield about a million times more energy per unit mass than do chemical reactions, so it's natural to try to get the energy that way.
NERVA got OK Isp (about a factor of 2 better than chemical rockets, something like 1000 seconds), but its thrust-to-weight ratio was pretty low, about 4 if I remember right. That's because it included a critical, operating nuclear reactor with an actively controlled chain reaction, and them thar things are heavy.
Thrust-to-weight is just as important as Isp to a rocket: higher thrust-to-weight means you can tote more fuel, payload, and structure for the same Isp, since you always have to have the mass of the engine itself around. By contrast to the NERVA's thrust-to-weight of about 4, the Space Shuttle main engines have a thrust-to-weight ratio of around 75. Since solid rockets are technically made out of their own fuel, their effective weight is much lower for this calculation (pretty much just the bell nozzle) and you might see numbers in the several-hundreds range.
Of course, one could always work on making the NERVA more lightweight -- but do you really want to optimize a nuclear reactor for mass, rather than safety? I didn't think so.
Now, for use in space, thrust-to-weight isn't so important. The rocket doesn't have to support itself against gravity, so low-mass engines that also produce low thrust are perfectly OK.
Of course, international treaty bans the use of critical nuclear reactors in space, but that alone wouldn't slow down our current administration very much.
[Nuclear reactors get flown into space all the time, but they always have much less than critical mass, relying on spontaneous decay to keep the chain reaction limping along at a constant rate. NERVA would require controlled reaction rates, hence a critical-mass reactor.]
The kinetic energy required to accelerate a gallon of gasoline to orbital speed is more than the chemical energy contained in the gasoline.
By contrast, "merely" lifting something up 100km doesn't require much energy at all.
So, er, no, leading-edge heat shields ought to be just fine. Fiberglass or carbon-fiber composites might even survive a flight or two without any shielding at all.
No, actually, I meant rm -rf .* -- that works pretty good on your home directory too, no matter where you run it :-)
Actually, rm -rf .* works considerably better -- it "sees" .. as well as your visible files.