I agree with your statement that the video iPod is inevitable.
It's also very interesting to note the following:
Go to www.apple.com/movies. "You don't have permission to access/movies on this server"
Go to www.apple.com/umptysquat. "Trying to find something at Apple?"
As Bill said from _Bill and Ted's Excellent Adventure_, "Ted... something strange is afoot at the Circle K."
However, and this is only my two cents, I don't think the technology / battery life / screen size / processor speed is quite there yet to show H.264 on a portable system in a marketable, affordable package. Give it two more years.
Not a single comment and the site is alread slahsdotted. sigh.
But this is an interesting trend: Apple has monopolized the headlines recently. ArsTechnica is all about Apple, Slashdot can't seem to get enough of them, and now Microsoft is emulating its Apple product?
What's next, Intel Processors branded with "Apple Outside" stickers on them?
Spoofed or not, the success rate of these hackers puts the Department of Defense to shame:
The number of attempted intrusions from all sources identified by the Pentagon last year totaled about 79,000, defense officials said, up from about 54,000 in 2003. Of those, hackers succeeded in gaining access to a Defense Department computer in about 1,300 cases. The vast majority of these instances involved what VanPutte called "low risk" computers.
I don't care if it's Low Risk; a 1.6% success rate is unacceptably high for Department of Defense computers.
In a similar vein, if I knew that 1.6% of all women in the world would sleep with me, I'd be a much happier camper.
This is a link to an old Slashdot story from maybe three years ago, that very eloquently talks about how the instantaneous nature of email, IM, and business in general these days is affecting people.
After reading it, I have turned off all notifications on my computer, and haven't looked back since. It's nice to be master of your own domain, even if it is a tiny one.
IAAAP (I Am Also A Physicist), and let me (humbly -- your explanation was really good) add some more meat to your description.
Physicists have a really, really hard time explaining *why* gravity is 10^42 times weaker than all other forces. (If you really want to split hairs, it's about 10^38 times weaker than the Weak Force, but what's an order of magnitude among friends?) Gravity appears to be a completely different manifestation than the electromagnetic, weak, and strong forces of nature. This irks many, and they try to rectify that by a Grand Unifying Theory (GUT).
One recent shot at explaining all this was well laid out in this article in Physics Today (subscription required, sorry) from 2002. In short, it theorized that gravity exists in 11 dimensions, not just 3, over short distances. Over some distance, the force known as gravity would "collapse" back down to our traditional 3. The fact that it acted over 11 dimensions, not 3, made gravity drop off as something like 1/r^10. This could help explain the apparent weakness of gravity.
IIRC, the authors predicted that gravity would get measurably stronger at small distances, as it was acting in many dimensions at once. Towards the upper end of their estimates, they predicted that gravity could be measurably stronger at distances around 3-5 millimeters.
As I read this latest discovery, it appears to throw water on that attempt to unify gravity with everything else. Back to the drawing board.
Was I the only one who read the subject line and read, "Who will buy Google next"?
Which could lead to a much more interesting discussion...
Is Micro$oft eager enough to buy it? Is it dumb enough to think it could? Or maybe a company like Oracle, that might be able to find a business case for purchasing a company like Google?
Or Apple... "Hey, we've improved Spotlight with our new Google technology!!"
To further the alarmist attitude, take this quote FTA:
Experts believe a crude plutonium bomb could be designed and assembled by terrorists possessing no greater level of skill than needed by the AUM cult to attack the Tokyo underground with nerve gas in 1995.
That is just plain wrong. The shape needed for a plutonium bomb to go off needs some very high tolerances, not achievable by average guys in a garage shop. Also throw in the fact that you must get the Plutonium-240 out of the PU-239, or it will prematurely detonate and all you're left with is a fizzle, not a true bomb. [The Pu-240 spontaneously fissions, and it will kick-start the fission process before the majority of the -239 kicks in, and it blows apart the sphere before you really get any significant neutron multiplication]
Perhaps the above fits in with the "crude plutonium bomb", but the way it's portrayed sure is alarmist.
Re:Is computational power the only thing missing?
on
Wormholes Unstable (BBC)
·
· Score: 3, Informative
Sigh. Okay, I'll bite.
Suppose, a few years from now, individual processors can do 100 trillion floating point operations per second. And you wire up 20,000 of these nodes in parallel. And suppose each floating point operation can magically operate one of those 10 to the 60th things-that-it-needs to (TFA didn't say *what* had to be controlled to within one in 10 to the 60th).
That's still 10^34 years. Not counting leap years.
As currently designed, nuclear engines in no way enable faster travel to the outer planets. They just allow a lot of energy packed into a small space.
Voyagers 1 and 2 made the trip to Jupiter in a handy two years. Galileo did it in a little over 3. Cassini took about 4.5 years to get to Jupiter.
As planned now, the Prometheus reactor, if one is ever sent to Jupiter, is not allowed to use gravity assists. This means it will take about 8 to 9 years given current ion propulsion thrusts to get it to Jupiter. Using a nuclear reactor to provide high voltages to spew xenon out the back end does not provide a whole lot of thrust, but it sure can be efficient.
The biggest advantage of putting nuclear reactors in space is the availability of lots of power on station for a fairly long period of time. It enables things like high-gain antennas, wide aperture arrays for surface mapping, or who knows what else -- if you build it, they will come. Give a scientist 200 kW (electric) to play with on Europa, I'm sure he'll find a good use for it.
Personally, I'm worried that SpaceShipTwo will never get off the ground:
Second System Syndrome.
I didn't come up with that term; it was first coined by Chuck Thacker of PARC. (great book for any engineer, by the way) You may come up with a great design for your original version, but often times the second version gets so bogged down with extra bells and whistles trying to be better than the first, that it never gets anywhere.
PARC suffered from SSS with their supercomputer, PARC did it again with the successor to the Alto -- the Star, and history has shown it again and again.
In other news, scientists have just heard what they believe to be a radio transmission from the long-lost Beagle 2 space probe. The radio transmission was very brief; it said only, "You're welcome. --Beagle 2."
Theories abound, but some crazy British supporters insist the Beagle2 is hiding behind a rock in this photo, the last photo taken from the Opportunity rover.
Yeah, that's great, but could somebody please knit me a solution to the Poincare Conjecture? It would really help me out. I promise to split the US $1 million for the prize money.
Alternatively, a Klein bottle would be neat, but it might take a while. Some of the stiches probably get kinda small....
Geosynchronous, at 22,500 miles out, is simply too far away to be viable.
The other issue at hand here is cost: the average life span of a spy satellite is about 3.5 or 4 years. The optics start to fog over due to radiation from the sun, and on something as precise as a spy satellite, that's a big deal. Plus space junk... my bet is that China launches 4 spy satellites in useless orbits by 2020 then gives up because it's just too darned expensive.
The maintenance on that kind of system would just be too darned expensive. Any GPS experts out there to lend credibility to this?
Ah, but this is being taken care of. It's just one of those things that are often done below the news reporter's radar screens. From this report, there is the following:
Currently, there are long lines of vessels waiting to be dismantled at the three facilities that have been receiving Western (especially U.S.) technical assistance: Nerpa (Murmansk), Zvezdochka (Severodvinsk), and the Zvezda Far Eastern Shipyard (Bolshoy Kamen). Until the technical and social problems at these facilities are resolved, the rates of dismantlement and use of Western equipment will remain well below peak efficiency. Current funds also do not provide for the long-term maintenance of the U.S. equipment, putting into question the issue of whether these yards will be able to continue working even at existing rates in the future, much less fulfill plans for accelerated submarine dismantlement. In addition, many of the submarines that need to be dismantled are located at shipyards without adequate facilities to do this work.
"Technical assistance" is provided by experienced engineers from Puget Sound Naval Shipyard, the same guys who are actively preventing this same situation from building up on the US's side.
The problem is, in the heyday of the Cold War, it was us vs. them, and we were each trying to pump out as many as possible. Money was spent like no object designing and building these things. We won, they lost, and now we have to help them foot the other HALF of the bill: paying for the dismantlement.
--The captain of another boat, upon hearing that Dana's captain wished to visit with him. The previous day, Dana's ship had accidentally rammed into the (anchored) boat.
First of all, congratulations. You're going to have a good time, as long as you understand that life is what you make of it. Don't expect hundreds of other slashdot geeks to come knocking on your dorm room door. Go out and meet people.
EverQuest does not count as 'going out and meeting people.'
As I recall, 90% of all laptop thefts occur because they get left unattended in libraries, cafeterias, study rooms, etc. Get yourself a nice shoulder bag for your laptop, and get used to the heft of having a laptop in it -- then you'll instantly know when it's not in there and you've forgotten in.
If you're paranoid about security in the dorm room, talk with your parents about renter's insurance. They can probably put a quick rider (or floater? IANAIS [insurance salseman]) on their homeowners' insurance for you and also cover your rented goods. Renters insurance, for the level of stuff you have, should be like $10 per month, I think. Be nice to your parents.
Lock your door, especially if people know you have the latest 256 MB nVidia QForce M-39 transponder video card in your ultraportable laptop.
It's like the 'hunt' for the neutrino, and scientists have been following that methodology ever since.
In 1930 or so, Wolfgang Pauli noticed that in all interactions, this strange combination of variables (what we now call spin) stayed constant through those interactions. But he couldn't fully explain beta-decay, or when the nucleus of an atom spits out an electron... this 'spin' wasn't conserved.
So, Pauli invents an incredible particle: it has little or no mass, hardly ever interacts with anything, but carries spin. It helped his equations balance.
Naturally, most of the scientific world scoffed at his idea at the time: it implied that hundreds of trillions of these things would be flying through space every second. AND they were undetectable?!? Quite a stretch.
But history bore him out, and neutrinos exist. You can see a history of the neutrino here, for more info, including current discrepancies with our understanding of neutrinos.
Quantum mechanics kinda developed the same way... crazy math with weird conclusions went AHEAD of experiments, and those experiments bore out the math 5 or 10 years later. I believe the same approach is being taken for the matter in the universe (WMAP predicitons), as well as the higgs boson.
This sounds more like vote-getting blabber than science-promoting talk.
One, somebody needs to pony up some serious dough for this. Moon mission would be, in my humble estimation, about $10B. The price tag is going to be much, much harder to swallow when the Big Bad Soviet Union isn't around to defeat.
Two, what about fixing the Space Shuttle? Project Prometheus? Making the ISS financially stable and properly crewed? There aren't the resources for this.
Don't get me wrong. I really, really want to see manned space flight get the heck out of lower earth orbit. But it's difficult to believe in an election year. It didn't work for his dad.....
I have an honest question, and maybe someone who's familiar with this kind of stuff can elucidate. IANARS.
Why not retro rocket thrusters? Okay, okay, I speculate they're heavy and they're complicated, but they seemed to work pretty well on the Vikings and the Russian-built Venera. And all the lunar landings, too. Can someone quantify *how* much heavier, or how much bulkier retro-rocket thrusters are than airbags? I really have no idea.
"There are no technical hurdles to producing a Powerbook G5. It could easily appear in January," Glaskowsky says.
I dunno... that sounds more like rumor-mill feedstock. 'No technical hurdles'? Seems far-fetched. Reducing the heat output is a good start, but the 970 still eats power. Something like 74W, IIRC. Most portable chips draw something in the 20W range. Again, reducing the transistor size is a good start, but there are significant hurdles still to be jumped if it's going to fit in a laptop with a reasonable battery life -- and not burn poor, unsuspecting scientists' winkies.
Billy Joel's "Songs in the Attic" was the first album ever released on CD, way back in 1981. One could argue that CD's didn't really catch on until 10 years later.
DVD's were supposed to flood the market for Christmas 1996, but didn't quite make it in time. That's right, kids, 1996. The titles were originally released in Japan: Blade Runner, The Assassin, The Fugitive, and Eraser. (yechhh) Then the first ones in the US were on March 19, 1997, and were IMAX remakes. Batman and Space Jam came next.
Seven years ago. I think it's fair to say DVDs have taken a strong hold in a pretty short period of time.
Microsoft has a lot more to do to present a "cool factor" than just get its hardware shown on popular TV shows in conspicuous places. Although there is probably an Apple rep on the site of "24" making sure that Apple logo is as blatantly obvious as possible, it wouldn't be the same with most of Microsoft's offerings.
It has to look cool, which, IMHO, Microsoft has not grasped yet. A large part of the appeal -- and probably a good reason why the directors of 24 allow it -- of Apple is the cool factor. The hardware is slick, the buttons are shiny, and it doesn't look like most other computers out there.
What product would microsoft have in its arsenal that could fill the above description?
I should think it would be exceedingly hard for a marketing community to market its 'immunity' to virii -- even a marketing staff as highly trained as whatever Apple hires -- without setting itself up as the next target.
Hypothetical advertisement: "Hey, we're Macs, and we don't have viruses."
I guarantee you that every virus writer and his(/her?) grandmother would flock to OS X and start writing viruses with reckless abandon. Apple, Linux, Amiga, Commodore 64, and whatever other less-used operating system is probably perfectly happy to have its users sitting fat, dumb, and happy and not bragging about it.
Back when my battery was new, I could get 5 hours out of the machine, if I used it only to play mp3's. In other words, the screen was off and the volume was pretty low (3 or 4).
Obviously, turn the screen as low as you can stand it. Quit all not-used programs. And if you're really a power miser, turn off all your network connections (AirPort, modem, ethernet, etc.)
Lastly, make sure the processor speed is set to "reduced" in the energy saver prefs. It'll add another 30-45 minutes to your battery life.
Overall, we're really only about 33% successful at it. Space Travel is Not Easy.
It's also very interesting to note the following:
Go to www.apple.com/movies. "You don't have permission to access /movies on this server"
Go to www.apple.com/umptysquat. "Trying to find something at Apple?"
As Bill said from _Bill and Ted's Excellent Adventure_, "Ted ... something strange is afoot at the Circle K."
However, and this is only my two cents, I don't think the technology / battery life / screen size / processor speed is quite there yet to show H.264 on a portable system in a marketable, affordable package. Give it two more years.
But this is an interesting trend: Apple has monopolized the headlines recently. ArsTechnica is all about Apple, Slashdot can't seem to get enough of them, and now Microsoft is emulating its Apple product?
What's next, Intel Processors branded with "Apple Outside" stickers on them?
The number of attempted intrusions from all sources identified by the Pentagon last year totaled about 79,000, defense officials said, up from about 54,000 in 2003. Of those, hackers succeeded in gaining access to a Defense Department computer in about 1,300 cases. The vast majority of these instances involved what VanPutte called "low risk" computers.
I don't care if it's Low Risk; a 1.6% success rate is unacceptably high for Department of Defense computers.
In a similar vein, if I knew that 1.6% of all women in the world would sleep with me, I'd be a much happier camper.
This is a link to an old Slashdot story from maybe three years ago, that very eloquently talks about how the instantaneous nature of email, IM, and business in general these days is affecting people.
After reading it, I have turned off all notifications on my computer, and haven't looked back since. It's nice to be master of your own domain, even if it is a tiny one.
Physicists have a really, really hard time explaining *why* gravity is 10^42 times weaker than all other forces. (If you really want to split hairs, it's about 10^38 times weaker than the Weak Force, but what's an order of magnitude among friends?) Gravity appears to be a completely different manifestation than the electromagnetic, weak, and strong forces of nature. This irks many, and they try to rectify that by a Grand Unifying Theory (GUT).
One recent shot at explaining all this was well laid out in this article in Physics Today (subscription required, sorry) from 2002. In short, it theorized that gravity exists in 11 dimensions, not just 3, over short distances. Over some distance, the force known as gravity would "collapse" back down to our traditional 3. The fact that it acted over 11 dimensions, not 3, made gravity drop off as something like 1/r^10. This could help explain the apparent weakness of gravity.
IIRC, the authors predicted that gravity would get measurably stronger at small distances, as it was acting in many dimensions at once. Towards the upper end of their estimates, they predicted that gravity could be measurably stronger at distances around 3-5 millimeters.
As I read this latest discovery, it appears to throw water on that attempt to unify gravity with everything else. Back to the drawing board.
Was I the only one who read the subject line and read, "Who will buy Google next"?
Which could lead to a much more interesting discussion ...
Is Micro$oft eager enough to buy it? Is it dumb enough to think it could? Or maybe a company like Oracle, that might be able to find a business case for purchasing a company like Google?
Or Apple ... "Hey, we've improved Spotlight with our new Google technology!!"
Experts believe a crude plutonium bomb could be designed and assembled by terrorists possessing no greater level of skill than needed by the AUM cult to attack the Tokyo underground with nerve gas in 1995.
That is just plain wrong. The shape needed for a plutonium bomb to go off needs some very high tolerances, not achievable by average guys in a garage shop. Also throw in the fact that you must get the Plutonium-240 out of the PU-239, or it will prematurely detonate and all you're left with is a fizzle, not a true bomb. [The Pu-240 spontaneously fissions, and it will kick-start the fission process before the majority of the -239 kicks in, and it blows apart the sphere before you really get any significant neutron multiplication]
Perhaps the above fits in with the "crude plutonium bomb", but the way it's portrayed sure is alarmist.
Suppose, a few years from now, individual processors can do 100 trillion floating point operations per second. And you wire up 20,000 of these nodes in parallel. And suppose each floating point operation can magically operate one of those 10 to the 60th things-that-it-needs to (TFA didn't say *what* had to be controlled to within one in 10 to the 60th).
That's still 10^34 years. Not counting leap years.
I'm not holding my breath ...
As currently designed, nuclear engines in no way enable faster travel to the outer planets. They just allow a lot of energy packed into a small space.
Voyagers 1 and 2 made the trip to Jupiter in a handy two years. Galileo did it in a little over 3. Cassini took about 4.5 years to get to Jupiter.
As planned now, the Prometheus reactor, if one is ever sent to Jupiter, is not allowed to use gravity assists. This means it will take about 8 to 9 years given current ion propulsion thrusts to get it to Jupiter. Using a nuclear reactor to provide high voltages to spew xenon out the back end does not provide a whole lot of thrust, but it sure can be efficient.
The biggest advantage of putting nuclear reactors in space is the availability of lots of power on station for a fairly long period of time. It enables things like high-gain antennas, wide aperture arrays for surface mapping, or who knows what else -- if you build it, they will come. Give a scientist 200 kW (electric) to play with on Europa, I'm sure he'll find a good use for it.
Second System Syndrome.
I didn't come up with that term; it was first coined by Chuck Thacker of PARC. (great book for any engineer, by the way) You may come up with a great design for your original version, but often times the second version gets so bogged down with extra bells and whistles trying to be better than the first, that it never gets anywhere.
PARC suffered from SSS with their supercomputer, PARC did it again with the successor to the Alto -- the Star, and history has shown it again and again.
Theories abound, but some crazy British supporters insist the Beagle2 is hiding behind a rock in this photo, the last photo taken from the Opportunity rover.
Alternatively, a Klein bottle would be neat, but it might take a while. Some of the stiches probably get kinda small ....
The other issue at hand here is cost: the average life span of a spy satellite is about 3.5 or 4 years. The optics start to fog over due to radiation from the sun, and on something as precise as a spy satellite, that's a big deal. Plus space junk ... my bet is that China launches 4 spy satellites in useless orbits by 2020 then gives up because it's just too darned expensive.
The maintenance on that kind of system would just be too darned expensive. Any GPS experts out there to lend credibility to this?
Currently, there are long lines of vessels waiting to be dismantled at the three facilities that have been receiving Western (especially U.S.) technical assistance: Nerpa (Murmansk), Zvezdochka (Severodvinsk), and the Zvezda Far Eastern Shipyard (Bolshoy Kamen). Until the technical and social problems at these facilities are resolved, the rates of dismantlement and use of Western equipment will remain well below peak efficiency. Current funds also do not provide for the long-term maintenance of the U.S. equipment, putting into question the issue of whether these yards will be able to continue working even at existing rates in the future, much less fulfill plans for accelerated submarine dismantlement. In addition, many of the submarines that need to be dismantled are located at shipyards without adequate facilities to do this work.
"Technical assistance" is provided by experienced engineers from Puget Sound Naval Shipyard, the same guys who are actively preventing this same situation from building up on the US's side.
The problem is, in the heyday of the Cold War, it was us vs. them, and we were each trying to pump out as many as possible. Money was spent like no object designing and building these things. We won, they lost, and now we have to help them foot the other HALF of the bill: paying for the dismantlement.
--A concerned reader.
"Oh? Did he bring his ship with him this time?"
--The captain of another boat, upon hearing that Dana's captain wished to visit with him. The previous day, Dana's ship had accidentally rammed into the (anchored) boat.
First of all, congratulations. You're going to have a good time, as long as you understand that life is what you make of it. Don't expect hundreds of other slashdot geeks to come knocking on your dorm room door. Go out and meet people.
EverQuest does not count as 'going out and meeting people.'
As I recall, 90% of all laptop thefts occur because they get left unattended in libraries, cafeterias, study rooms, etc. Get yourself a nice shoulder bag for your laptop, and get used to the heft of having a laptop in it -- then you'll instantly know when it's not in there and you've forgotten in.
If you're paranoid about security in the dorm room, talk with your parents about renter's insurance. They can probably put a quick rider (or floater? IANAIS [insurance salseman]) on their homeowners' insurance for you and also cover your rented goods. Renters insurance, for the level of stuff you have, should be like $10 per month, I think. Be nice to your parents.
Lock your door, especially if people know you have the latest 256 MB nVidia QForce M-39 transponder video card in your ultraportable laptop.
Again, good luck.
In 1930 or so, Wolfgang Pauli noticed that in all interactions, this strange combination of variables (what we now call spin) stayed constant through those interactions. But he couldn't fully explain beta-decay, or when the nucleus of an atom spits out an electron ... this 'spin' wasn't conserved.
So, Pauli invents an incredible particle: it has little or no mass, hardly ever interacts with anything, but carries spin. It helped his equations balance.
Naturally, most of the scientific world scoffed at his idea at the time: it implied that hundreds of trillions of these things would be flying through space every second. AND they were undetectable?!? Quite a stretch.
But history bore him out, and neutrinos exist. You can see a history of the neutrino here, for more info, including current discrepancies with our understanding of neutrinos.
Quantum mechanics kinda developed the same way ... crazy math with weird conclusions went AHEAD of experiments, and those experiments bore out the math 5 or 10 years later. I believe the same approach is being taken for the matter in the universe (WMAP predicitons), as well as the higgs boson.
Just my 0.02 euro.
One, somebody needs to pony up some serious dough for this. Moon mission would be, in my humble estimation, about $10B. The price tag is going to be much, much harder to swallow when the Big Bad Soviet Union isn't around to defeat.
Two, what about fixing the Space Shuttle? Project Prometheus? Making the ISS financially stable and properly crewed? There aren't the resources for this.
Don't get me wrong. I really, really want to see manned space flight get the heck out of lower earth orbit. But it's difficult to believe in an election year. It didn't work for his dad.....
Why not retro rocket thrusters? Okay, okay, I speculate they're heavy and they're complicated, but they seemed to work pretty well on the Vikings and the Russian-built Venera. And all the lunar landings, too. Can someone quantify *how* much heavier, or how much bulkier retro-rocket thrusters are than airbags? I really have no idea.
"There are no technical hurdles to producing a Powerbook G5. It could easily appear in January," Glaskowsky says.
I dunno ... that sounds more like rumor-mill feedstock. 'No technical hurdles'? Seems far-fetched. Reducing the heat output is a good start, but the 970 still eats power. Something like 74W, IIRC. Most portable chips draw something in the 20W range. Again, reducing the transistor size is a good start, but there are significant hurdles still to be jumped if it's going to fit in a laptop with a reasonable battery life -- and not burn poor, unsuspecting scientists' winkies.
Billy Joel's "Songs in the Attic" was the first album ever released on CD, way back in 1981. One could argue that CD's didn't really catch on until 10 years later.
DVD's were supposed to flood the market for Christmas 1996, but didn't quite make it in time. That's right, kids, 1996. The titles were originally released in Japan: Blade Runner, The Assassin, The Fugitive, and Eraser. (yechhh) Then the first ones in the US were on March 19, 1997, and were IMAX remakes. Batman and Space Jam came next.
Seven years ago. I think it's fair to say DVDs have taken a strong hold in a pretty short period of time.
It has to look cool, which, IMHO, Microsoft has not grasped yet. A large part of the appeal -- and probably a good reason why the directors of 24 allow it -- of Apple is the cool factor. The hardware is slick, the buttons are shiny, and it doesn't look like most other computers out there.
What product would microsoft have in its arsenal that could fill the above description?
Hypothetical advertisement: "Hey, we're Macs, and we don't have viruses."
I guarantee you that every virus writer and his(/her?) grandmother would flock to OS X and start writing viruses with reckless abandon. Apple, Linux, Amiga, Commodore 64, and whatever other less-used operating system is probably perfectly happy to have its users sitting fat, dumb, and happy and not bragging about it.
Obviously, turn the screen as low as you can stand it. Quit all not-used programs. And if you're really a power miser, turn off all your network connections (AirPort, modem, ethernet, etc.)
Lastly, make sure the processor speed is set to "reduced" in the energy saver prefs. It'll add another 30-45 minutes to your battery life.