Microsoft Research put a web server on a mobile phone several years ago, back when I still worked for them.
A cow-orker (or should that be core-searcher? named Kai Rannenburg did the dirty deed. Kai left MSR shortly before I did. Chase down to Kai's exp-rojects at http://research.microsoft.com/security/ for more information.
So you were using the NVidia driver? These days the NVidia driver installer automatically sets the X display driver to 'nvidia' so you don't have to do that anymore.
I've no reason to doubt the correctness of your statement.
This machine was purchased from Dell about a month ago. Dell pre-installed RedHat for me. Dell are well known to build to order, rather than keeping a large stock. Therefore, I deduce that Dell installed the OS and X about a month ago.
From this chain of reasoning, I deduce that "these days" refers to the last few weeks.
What's your video card? I can't even remember the last time I had to manually configure X. Since 4-5 years ago practically all Linux distro installers began to autodetect video card settings and automatically configure X for you at install time.
Real story. We're mostly a Sloaris shop at work but I got myself a Linux box from Dell for two reasons --- it was roughly half the price and twice the cpu performance of a SunBlade that everyone else uses, and I have enough experience with both environments to be able to work around any incompatibilities between my system and the rest of the lab's. My purchase was made with an explicit goal of trialling a switch over at the next major rekitting exercise
The Precision 470 rolled up, along with its spiffy 2001FP display with native resolution 1600x1200. (Not as spiffy as our Suns, btw, which run at 1920x1280.) I could not get the damn thing to run X at a resolution greater than 1280x1024, with predictably cruddy results. Everything would be sharp and clear until X came up, so the video card and the display themselves were clearly working perfectly.
After much grubbing around in config files, rummaging in Google, and the like, I eventually discovered that X was using a vesa driver and that I had to replace it with the magic word "nvidia".
I have never come across a Windows system shipped in a comparable state. They have invariably worked straight out of the box.
The largest nuclear bombs detonated by humans have released an energy of approximately 400 quadrillion joules. This is about 20,000,000 times the energy expended by a Saturn V rocket, one of humanity's most impressive feats of engineering.
Something is wrong there.
A Saturn V was loaded with a few kilotons of chemical explosives. The largest thermonuclear explosion was a few tens of megatons TNT equivalent. To a good enough approximation, all chemical explosives are about as powerful as TNT.
I suggest that you mean 20 thousand times the energy, not 20 million.
AFAIK, the source for this quote is Tony Sale, the guy who lead the Colossus-rebuild project at BP. He certainly claimed in my presence that Colossus was about ten times faster than the program he'd written for Pentium PC (not P4). This would be in summer 1998, when Colossus was just about functional again.
The rebuilt Colossus is an interesting machine, BTW, and well worth going to see. If you get chance to get up real close, don't go poking fingers in it. There's 400V DC on uninsulated conductors.
The parent poster is correct: a properly used Enigma machine is effectively unbreakable with the technology of the day and, for that matter, the technology of the next few decades too.
The majority of the users of the Enigma machine were not using it properly and so left cracks for BP to exploit. All this is well documented by people who do know a great deal about cryptographic systems. Some of them worked for BP and have in-depth first hand knowledge of what they write about.
Even today's technology, that in the open literature anyway, still has real difficulties breaking Naval Enigma without the weaknesses introduced by the German users of the system. Read the site carefully and you will discover that important amounts of key material are already known, thereby greatly reducing the amount of computation required to find the rest of the key. And even with this assistance, approximately a cpu year is needed to break the encryption.
All this strongly suggests that Naval Enigma isn't a bad cryptosystem and certainly a good one for the day. There have been many much worse ones fielded in recent years.
The hardware of the day, the 8086 and 8088 with 64-256K RAM, could not sensibly run the Unix of the day. Almost total non-existent memory management for one thing.
That almost forced a CP/M-like OS to be used. Thereafter, the great god of backwards compatibility ensured the evolution we saw.
Only relatively recently have we seen a migration to a fundamentally different kernel. Hiring the people from DEC inevitably made the new OS more similar to VMS than Unix.
Apple had to make a similar transition, for essentially the same reasons, but they were unable to hire good people sufficient familiar with anything but Unix. As a result, the Mac now runs a bastardized BSD.
they used to be the empire where the sun never set...
Check it out --- the sun still never sets on the British Empire. It's a damned close run thing, though, since the loss of Hong Kong.
IIRC, the worst case overlap between the sun setting in Pitcairn and rising over Diego Garcia and others in the British Indian Ocean Territories is something like 37 minutes.
We had instant start 20 years ago with the Sinclair Spectrum and the Atari ST. I've been waiting two decades for PCs to catch up;-)
My HP Jornada 820 is instant-on. It's a WinCE 2.1 machine in subnotebook format. Real keyboard, VGA display, modem, USB for external mouse, trackpad, CF and PCMCIA slots (so ethernet is installed in the first and 128M storage lives in the other), OS, IE, Word, Access, Excel, Outlook Express, Powerpoint all in ROM. Ten to twelve hours battery life.
It's a really nice machine that I take on trips in preference to "real" laptops. Never understood why HP stopped making them and other manufacturers never started.
Weight. Even the components of a big reactor are heavy. A Tokamak style fusion reactor that exhausted into the engine feed would be the best system, and probably have the thrust to take us to the stars (1 G acceleration for 1 year to get damn close to lightspeed).
Fuel mass, especially for your interstellar craft.
Assume that the craft does indeed get "damn close to lightspeed" and let's completely ignore the problems of running into inter{planetary,stellar} {dust,gas} at that velocity. From good old E=mc^2, we know that the craft will increase in mass substantially as seen by an Earth-bound observer. That mass has to come from somewhere, and that has to be from the fuel. A fusion reactor is at most 1% efficient, so even under ideal circumstances the payload has to be less than 1% of the launch mass if it is to get anywhere near lightspeed.
The 1970's Daedalus starship design from the British Interplanetary Society went into quite some detail. Their design used D-He3 fusion, 50 thousand tonnes of propellant burned for about a year and accelerated a 500-tonne payload to about c/8 --- enough for a fly-by mission to Barnard's star in a few decades.
Yes, molecular oxygen does leave a mark in the solar spectrum as measured on Earth's surface, in fact. But that's because there is a LOT of oxygen lying between us and space. In lower concentrations, it becomes pretty dicey. Ozone, however, gives a nice, strong absorption and gives away the presence of O_2. So it has always been considered a preferable way to seeking out atmospheric oxygen.
I think we're in violent agreement. I claimed it is easy to see the O_2 absorption spectrum with a path length of around one atmosphere-meter; you that a couple of atmosphere-kilometres is sufficient
My point should be read in the article's context of characterizing extrasolar planetary atmospheres. It's a fair bet that if an ecosystem uses an oxygen ecomony like ours (catalytic photolysis of water and oxidation with free oxygen to complete the cycle) such a planet would also have a moderately high concentration of free O_2 and path lengths of a few kilometres should be available for light reflected from the planetary surface.
I don't doubt that O_3 is easier to detect in the near UV than O_2 is in the near-IR, but if spectroscopy of starlight reflected from an extrasolar terrestial planet is going to be done at all, it is likely to be possible to detect numerous molecules, including O_2, H_20, CO_2, N_H3 and CH_4 --- all in the appropriate regions of the EM-spectrum of course! Picking up other atmospheric species, such as N_2 and Ar, is likely to be rather difficult though their presence may perhaps be inferred from other properties of the atmosphere.
Oxygen, alas, leaves little mark in the spectrum since it's a homonuclear diatomic molecule and light tends to ignore it.
Well, yes and no. I was a molecular spectroscopist in an earlier life and found that the spectrum of O_2 was actually quite easy to detect in the deep red and near infrared though, to be fair, I had plenty of light available and was looking through several metres of O_2 at around 20% of an atmosphere partial pressure. The O_2 was a pain --- I was actually interested in the spectrum of gaseous CeO (cerium monoxide).
O_2 is indeed a homonuclear diatomic molecule, but a rather unusual one. It has a strong magnetic dipole transition in the near IR (high-lying vibrational levels creep into the deep red, as already mentioned). It's this transition which gives LOX its pretty pale blue colour.
I could go into much more detail but anyone interested will find at least as much information as they want by searching the regular places for stuff on molecular spectrosocopy.
Since the model assumes that a galaxy is a fluid (on a large scale), the model would predict fluid-like phenomena. What I wonder is if there is a galactic analogue to solitary waves. How would these manifest? (A friend wrote his thesis on solitons)
Yes there are analogues and easily visible manifestations are spiral arms.
Treating galaxies as fluids has been done for many years and the models have been quite successful. I think it was found back in the 70's that spiral arms could be modelled rather well as density waves in a rotating disc of fluid.
Gosh. So, if Microsoft realised that, why then they'd naturally oppose software patents. You see, they'd realise that if software patents were outlawed, then they wouldn't need the expanse of maintaining a large patent portfolio to defend themselves, and they'd be free of the threat of Eolas type lawsuits into the bargain.
Microsoft, like all corporations have to live with the things way are and not the way they'd like them to be. That doesn't mean they like the status quo, nor does it mean they are not working to change things. Note that I am most emphatically not saying that MS specifically does not want patents, only that patents exist and MS, like all corporations, has to take them into account.
In our present environment, patents are a fact of life. MS and its competitors have evolved to take that situation into account.
Frankly, if you aren't willing/able to spend the time to look, just go ahead and use it.
Unintentional GPL violations just require you to remove the GPL code from your product. As long as you aren't sitting there trying to be evil, and you've made a good faith effort to only use LGPL stuff, I can't imagine any judge would seriously go after you.
I know the FSF doesn't. 100% of the time, when the FSF goes after someone, their *first* request is to settle the dispute by having the offending party remove the offending code.
You are saying that either you avoid the GPL and write your own code from the start, or violate the GPL and write your own code when you get caught.
Both of these approaches cost money, because both of them involve writing your own code. Why, then, use GPL code? Not only does it cost you the same development effort as rolling your own, it also causes disruption to the business part way through the life of the product (something, incidentally, which is likely to be much more expensive than the code development) it also causes embarassment to the company and does not improve the quality of their employees' resumes.
IBM created the concept of open architecture with the first PC.
That is a bit of an exaggeration. IBM did not create the PC as an open architecture. The BIOS was an essential part of the architecture and IBM made it very clear that they held the copyright on that code. Compaq reverse-engineered the IBM BIOS in a well-documented clean room operation.
In my opinion, IBM and Compaq jointly made it simple for other companies to build PC-compatible machines.
Of course, the open architecture phenomenon goes back to long before the PC. The old-timers here will know who the BUNCH were and why they are relevant to this discussion.
In fact, tidal forces at decent proximity to the event horizon will make protons and neutrons flow like water (the quarks and gluons in some cases as well...)
Only for sufficiently small black holes. There is nothing locally special about a region near the event horizon. If the BH is big enough, 100M solar masses say, the tidal forces at its event horizon are small enough to let an astronaut pass through it without ill-effect. The same can not be said of the experiences encountered much closer to the singularity.
"Every law firm worth its salt has double-redundancy: old systems are kept and paper copies of all documents."
I can't tell if you're trying to say they're smart for keeping all these old systems around, or dumb for not using a standardized open format so that they wouldn't have to.
A paper copy of a document is in a standardized open format. Further, it is known to be readable for millenia if kept under the correct conditions. We're still able to read legal correspondence written 3500 years ago. Ok, that was written on papyrus but there really isn't all that much difference between papyrus and modern paper.
Which is why Starwisp, the closest thing we've got to a decent design for an interstellar probe, will accelerate at 115g.
Another design is now over 25 year old --- Daedalus. A Google search will tell you much more about it, but it's basically a very big rocket.
The reference design accelerates a 500 tonne payload to around 0.12c using a two stage rocket. The acceleration is around 0.1g and the rocket runs continuously for something over a year. When the probe reaches the target star (the reference mission was to Barnard's star just under 6 light years away) the rocket nozzle is used as the comms antenna.
Needless to say, this is a big rocket. It's a pulse fusion design, using 20 thousand tonnes of deuterium and 30 thousand tonnes of helium-3 for fuel. The only credible places to get that amount of He-3 is from the atmospheres of the gas giants. Assuming that the Daedalus probe was built in Jupiter orbit, the rocket exhaust would still be a spectacular sight even from earth.
Oh, interesting -- how come? The difficulty in working out the share of everyone involved in the effort?
Partly that. Partly dealing with foreign currency exchanges, bank charges, differing taxation regimes, corporate and university policies, postal charges to get chec{k,que}s out to participants,...
So, a puzzling question is why did the team take on RSA-200 rather than RSA-640?
Because dealing with prize money causes acute rectal discomfort and is more trouble than it's worth.
Although I played no role in this factorization, I've participated in several others, including RSA-129 (which got the whole business rolling) and RSA-576 with Kleinjung et al. In every case, the prize money was a disincentive.
I saw the movie a few hours ago. Parts of it were excellent. Parts of it were not. In particular, cutting the dialogue between Arthur and Mr Prosser was unforgivable, as was chopping out almost all the poetry reading and the removal of the "men were real men, women were real women and small furry creatures from Alpha Centauri were real small creatures from Alpha Centauri" elaboration of the business model of Magrathea.
Slashdot Mirror
Nope, not an oxymoron on this occasion.
Microsoft Research put a web server on a mobile phone several years ago, back when I still worked for them.
A cow-orker (or should that be core-searcher? named Kai Rannenburg did the dirty deed. Kai left MSR shortly before I did. Chase down to Kai's exp-rojects at http://research.microsoft.com/security/ for more information.
Paul
So you were using the NVidia driver? These days the NVidia driver installer automatically sets the X display driver to 'nvidia' so you don't have to do that anymore.
I've no reason to doubt the correctness of your statement.
This machine was purchased from Dell about a month ago. Dell pre-installed RedHat for me. Dell are well known to build to order, rather than keeping a large stock. Therefore, I deduce that Dell installed the OS and X about a month ago.
From this chain of reasoning, I deduce that "these days" refers to the last few weeks.
Paul
What's your video card? I can't even remember the last time I had to manually configure X. Since 4-5 years ago practically all Linux distro installers began to autodetect video card settings and automatically configure X for you at install time.
Real story. We're mostly a Sloaris shop at work but I got myself a Linux box from Dell for two reasons --- it was roughly half the price and twice the cpu performance of a SunBlade that everyone else uses, and I have enough experience with both environments to be able to work around any incompatibilities between my system and the rest of the lab's. My purchase was made with an explicit goal of trialling a switch over at the next major rekitting exercise
The Precision 470 rolled up, along with its spiffy 2001FP display with native resolution 1600x1200. (Not as spiffy as our Suns, btw, which run at 1920x1280.) I could not get the damn thing to run X at a resolution greater than 1280x1024, with predictably cruddy results. Everything would be sharp and clear until X came up, so the video card and the display themselves were clearly working perfectly.
After much grubbing around in config files, rummaging in Google, and the like, I eventually discovered that X was using a vesa driver and that I had to replace it with the magic word "nvidia".
I have never come across a Windows system shipped in a comparable state. They have invariably worked straight out of the box.
Paul
The largest nuclear bombs detonated by humans have released an energy of approximately 400 quadrillion joules. This is about 20,000,000 times the energy expended by a Saturn V rocket, one of humanity's most impressive feats of engineering.
Something is wrong there.
A Saturn V was loaded with a few kilotons of chemical explosives. The largest thermonuclear explosion was a few tens of megatons TNT equivalent. To a good enough approximation, all chemical explosives are about as powerful as TNT.I suggest that you mean 20 thousand times the energy, not 20 million.
Paul
AFAIK, the source for this quote is Tony Sale, the guy who lead the Colossus-rebuild project at BP. He certainly claimed in my presence that Colossus was about ten times faster than the program he'd written for Pentium PC (not P4). This would be in summer 1998, when Colossus was just about functional again.
The rebuilt Colossus is an interesting machine, BTW, and well worth going to see. If you get chance to get up real close, don't go poking fingers in it. There's 400V DC on uninsulated conductors.
Paul
The parent poster is correct: a properly used Enigma machine is effectively unbreakable with the technology of the day and, for that matter, the technology of the next few decades too.
The majority of the users of the Enigma machine were not using it properly and so left cracks for BP to exploit. All this is well documented by people who do know a great deal about cryptographic systems. Some of them worked for BP and have in-depth first hand knowledge of what they write about.
Even today's technology, that in the open literature anyway, still has real difficulties breaking Naval Enigma without the weaknesses introduced by the German users of the system. Read the site carefully and you will discover that important amounts of key material are already known, thereby greatly reducing the amount of computation required to find the rest of the key. And even with this assistance, approximately a cpu year is needed to break the encryption.
All this strongly suggests that Naval Enigma isn't a bad cryptosystem and certainly a good one for the day. There have been many much worse ones fielded in recent years.
Paul
If you paid as much as NASA did for the Saturn comms, you'd get pretty good cellphone connections too.
By and large, you get what you pay for
Paul
The hardware of the day, the 8086 and 8088 with 64-256K RAM, could not sensibly run the Unix of the day. Almost total non-existent memory management for one thing.
That almost forced a CP/M-like OS to be used. Thereafter, the great god of backwards compatibility ensured the evolution we saw.
Only relatively recently have we seen a migration to a fundamentally different kernel. Hiring the people from DEC inevitably made the new OS more similar to VMS than Unix.
Apple had to make a similar transition, for essentially the same reasons, but they were unable to hire good people sufficient familiar with anything but Unix. As a result, the Mac now runs a bastardized BSD.
Paul
Paul
Presumably they mean bloated and likely to suffer a premature death.
Paul
Check it out --- the sun still never sets on the British Empire. It's a damned close run thing, though, since the loss of Hong Kong.
IIRC, the worst case overlap between the sun setting in Pitcairn and rising over Diego Garcia and others in the British Indian Ocean Territories is something like 37 minutes.Paul
My HP Jornada 820 is instant-on. It's a WinCE 2.1 machine in subnotebook format. Real keyboard, VGA display, modem, USB for external mouse, trackpad, CF and PCMCIA slots (so ethernet is installed in the first and 128M storage lives in the other), OS, IE, Word, Access, Excel, Outlook Express, Powerpoint all in ROM. Ten to twelve hours battery life.
It's a really nice machine that I take on trips in preference to "real" laptops. Never understood why HP stopped making them and other manufacturers never started.
Paul
Fuel mass, especially for your interstellar craft.
Assume that the craft does indeed get "damn close to lightspeed" and let's completely ignore the problems of running into inter{planetary,stellar} {dust,gas} at that velocity. From good old E=mc^2, we know that the craft will increase in mass substantially as seen by an Earth-bound observer. That mass has to come from somewhere, and that has to be from the fuel. A fusion reactor is at most 1% efficient, so even under ideal circumstances the payload has to be less than 1% of the launch mass if it is to get anywhere near lightspeed.
The 1970's Daedalus starship design from the British Interplanetary Society went into quite some detail. Their design used D-He3 fusion, 50 thousand tonnes of propellant burned for about a year and accelerated a 500-tonne payload to about c/8 --- enough for a fly-by mission to Barnard's star in a few decades.
Paul
I think we're in violent agreement. I claimed it is easy to see the O_2 absorption spectrum with a path length of around one atmosphere-meter; you that a couple of atmosphere-kilometres is sufficient
My point should be read in the article's context of characterizing extrasolar planetary atmospheres. It's a fair bet that if an ecosystem uses an oxygen ecomony like ours (catalytic photolysis of water and oxidation with free oxygen to complete the cycle) such a planet would also have a moderately high concentration of free O_2 and path lengths of a few kilometres should be available for light reflected from the planetary surface.
I don't doubt that O_3 is easier to detect in the near UV than O_2 is in the near-IR, but if spectroscopy of starlight reflected from an extrasolar terrestial planet is going to be done at all, it is likely to be possible to detect numerous molecules, including O_2, H_20, CO_2, N_H3 and CH_4 --- all in the appropriate regions of the EM-spectrum of course! Picking up other atmospheric species, such as N_2 and Ar, is likely to be rather difficult though their presence may perhaps be inferred from other properties of the atmosphere.
Paul
Well, yes and no. I was a molecular spectroscopist in an earlier life and found that the spectrum of O_2 was actually quite easy to detect in the deep red and near infrared though, to be fair, I had plenty of light available and was looking through several metres of O_2 at around 20% of an atmosphere partial pressure. The O_2 was a pain --- I was actually interested in the spectrum of gaseous CeO (cerium monoxide).
O_2 is indeed a homonuclear diatomic molecule, but a rather unusual one. It has a strong magnetic dipole transition in the near IR (high-lying vibrational levels creep into the deep red, as already mentioned). It's this transition which gives LOX its pretty pale blue colour.
I could go into much more detail but anyone interested will find at least as much information as they want by searching the regular places for stuff on molecular spectrosocopy.
Paul
Yes there are analogues and easily visible manifestations are spiral arms.
Treating galaxies as fluids has been done for many years and the models have been quite successful. I think it was found back in the 70's that spiral arms could be modelled rather well as density waves in a rotating disc of fluid.
Paul
Microsoft, like all corporations have to live with the things way are and not the way they'd like them to be. That doesn't mean they like the status quo, nor does it mean they are not working to change things. Note that I am most emphatically not saying that MS specifically does not want patents, only that patents exist and MS, like all corporations, has to take them into account.
In our present environment, patents are a fact of life. MS and its competitors have evolved to take that situation into account.
Paul
Unintentional GPL violations just require you to remove the GPL code from your product. As long as you aren't sitting there trying to be evil, and you've made a good faith effort to only use LGPL stuff, I can't imagine any judge would seriously go after you.
I know the FSF doesn't. 100% of the time, when the FSF goes after someone, their *first* request is to settle the dispute by having the offending party remove the offending code.
You are saying that either you avoid the GPL and write your own code from the start, or violate the GPL and write your own code when you get caught.
Both of these approaches cost money, because both of them involve writing your own code. Why, then, use GPL code? Not only does it cost you the same development effort as rolling your own, it also causes disruption to the business part way through the life of the product (something, incidentally, which is likely to be much more expensive than the code development) it also causes embarassment to the company and does not improve the quality of their employees' resumes.Paul
That is a bit of an exaggeration. IBM did not create the PC as an open architecture. The BIOS was an essential part of the architecture and IBM made it very clear that they held the copyright on that code. Compaq reverse-engineered the IBM BIOS in a well-documented clean room operation.
In my opinion, IBM and Compaq jointly made it simple for other companies to build PC-compatible machines.
Of course, the open architecture phenomenon goes back to long before the PC. The old-timers here will know who the BUNCH were and why they are relevant to this discussion.
Paul
Only for sufficiently small black holes. There is nothing locally special about a region near the event horizon. If the BH is big enough, 100M solar masses say, the tidal forces at its event horizon are small enough to let an astronaut pass through it without ill-effect. The same can not be said of the experiences encountered much closer to the singularity.
Paul
I can't tell if you're trying to say they're smart for keeping all these old systems around, or dumb for not using a standardized open format so that they wouldn't have to.
A paper copy of a document is in a standardized open format. Further, it is known to be readable for millenia if kept under the correct conditions. We're still able to read legal correspondence written 3500 years ago. Ok, that was written on papyrus but there really isn't all that much difference between papyrus and modern paper.
Paul
Another design is now over 25 year old --- Daedalus. A Google search will tell you much more about it, but it's basically a very big rocket.
The reference design accelerates a 500 tonne payload to around 0.12c using a two stage rocket. The acceleration is around 0.1g and the rocket runs continuously for something over a year. When the probe reaches the target star (the reference mission was to Barnard's star just under 6 light years away) the rocket nozzle is used as the comms antenna.
Needless to say, this is a big rocket. It's a pulse fusion design, using 20 thousand tonnes of deuterium and 30 thousand tonnes of helium-3 for fuel. The only credible places to get that amount of He-3 is from the atmospheres of the gas giants. Assuming that the Daedalus probe was built in Jupiter orbit, the rocket exhaust would still be a spectacular sight even from earth.
Paul
Partly that. Partly dealing with foreign currency exchanges, bank charges, differing taxation regimes, corporate and university policies, postal charges to get chec{k,que}s out to participants, ...
Paul
Because dealing with prize money causes acute rectal discomfort and is more trouble than it's worth.
Although I played no role in this factorization, I've participated in several others, including RSA-129 (which got the whole business rolling) and RSA-576 with Kleinjung et al. In every case, the prize money was a disincentive.
Paul
ITYM "Beware of the leopard". HTH, HAND.
I saw the movie a few hours ago. Parts of it were excellent. Parts of it were not. In particular, cutting the dialogue between Arthur and Mr Prosser was unforgivable, as was chopping out almost all the poetry reading and the removal of the "men were real men, women were real women and small furry creatures from Alpha Centauri were real small creatures from Alpha Centauri" elaboration of the business model of Magrathea.
Paul