I remember that, during my teens, I once woke up in a hospital after a party. The doctor, with an extremely serious face, explained to my half-conscious brain that I had too little blood in my alcoholstream and that the condition had been corrected.
They could easily subsidize the hardware required to run the system. Airlines would even demand they pay for the extra fuel the heavier computers would require;-)
They could also go the thin client with beefy server route. Maybe not as good for a in flight entertainment system, but good enough for browsing and emailing.
But they will probably wait until people (read: their astroturf teams) start complaining OpenOffice is not Office and refuses to read their MSOOXML files before they announce their move that will "save" the airlines from the headaches of the previous system.
Never underestimate neither their weaseliness nor the depth of their pockets.
"linux is best in computers or embedded devices where you need high reliability and you want to be able to specify the exact amount of the functionality it should have."
While I would like to point out this is not about critical flight control systems (where I doubt any Linux would be certified as it costs a lot to be) and in-flight entertainment machines are OK to crash sometimes, the specific functionality is, probably, a win for Linux distros.
But, in the end, I suspect the real deal here is about price. The cheapest solution won. It would be hideously expensive to have Windows Vista PCs with Office 2007 on every seat of a jetliner.
For multi-threading apps, instead of multiple cores (nothing except caches are shared between complete CPU cores), it makes a lot of sense to have an HT-like architecture (multiple context stores, shared elements) that reduces the time it takes to do a context-switch. It would also help a lot to have a context-aware cache system where a swapped-in context would not wake up having to read every instruction from main memory.
Since not all threads will be runnable at any given time, having more cores instead of bigger caches could hurt performance instead of helping it. A bunch of context stores and a couple independent CPU elements will not give you the same performance boost than a second core will, but they take up a lot less silicon and may spare you many trips to main memory.
The IIgs was too little, too late to save the Apple II line. It came out after the Mac, the Amiga 1000 and the Atari 520ST. At the time, it ran mostly 8-bit Apple II software (so, it was an overexpensive IIe that could run few software the older brother couldn't) and the full GS/OS only appeared much later. As things happened, it stood little to no chance of survival. In the end, the Apple IIe outlived it.
If you ask me, what doomed the II line was the failure of the IIx project. Had they finished it as a next-gen Apple II and launched it before the competition, they would have ensured the survival of the II line for a couple more years.
Not to say I don't want one. Really, I want all three in my interesting computer collection.
The II (6502) to Lisa/Mac (68K) was not a real transition. They had different lines of computers with different software. The Lisa or the Mac was never compatible with the II (without an expansion card). Even the Apple III, which used the same 6502, was barely compatible with the II.
The 68K to PPC transition was very smooth and much of the software base ran without a hitch. It worked so well only 8.6 was fully PPC and I bet a lot of it still ran under 68K emulation.
The PPC to x86 transition was interesting. Seemingly nothing changed, but the computers got a whole lot faster.
They would need a decent FPU/vector accelerator for this kind of application. I say it's very feasible, but, perhaps, not the best way to go.
OTOH, it would be much more clever to use MIPS or ARMs as general-purpose CPUs. On the same space, it should be feasible to cram a couple ARM11s, lots of cache and a couple DSP/SPU/GPU-like cores sitting around the main cores. It would no longer be as low-power as these x86s, but it would be a lot more interesting for desktop workstations. I find the Tilera 64-core chip very interesting in this regard. If they could build 32 CPU/FPU pairs instead of 64 FPU-less cores, it would be even more impressive. I would love to see a desktop made with them.
Best of all, they would be remarkably Windows-proof.
Not being properly educated is a sure way to thing something you don't quite understand is crap.
Big Bang is still a very good theory, but it's a theory. People will adjust it to match the observed facts and, when it becomes increasingly hard to adjust it, people will prefer other better theories that may fit the facts with less tweaking.
And no. "God and angels" is by no means a decent explanation.
"Apollo program got started by replacing the warheads with men in tin can"
I think that would be Mercury. All Apollo vehicles were ridiculously large for ICBMs.
"Or take shuttle. It was designed on paper, and the very first hardware iteration was declared operational configuration. Thats just nuts."
That happens when politics and engineering get too close. The shuttle has so many different mission objectives it could not (at this time with present technology) be a good solution.
As for this evolutionary approach, you need economic incentive. Passenger flights evolved because there were passengers willing to pay, airlines willing to buy and operate planes and, thus, companies willing to build them. There is no such thing for passenger spaceflight.
There _is_ for unmanned launchers and you see all kinds of evolution there. From the early launchers to Ariane V there is a lot between. There is little less variation than with airplanes mainly because the problem is a much more difficult one and there are fewer possible solutions with the same technologies. Right now, we can't make anti-gravity or warp-drive, so, we will have to stick to chemical rockets. And nuclear-thermal is a regulation nightmare. As much as I love the idea of cheap and widespread space-travel, I don't want people to have easy access to nuclear fuel.
If I would imagine a future with cheap access to space, I would start with commercial sub-orbital flight. If you figure out a way to reliably fly 60 passengers between China and California for a reasonable fare in a couple hours and without complex infrastructure at the ends, I think you would get filthy rich. Even if you could do so with cargo, DHL and FedEx would order everything you could build. If it does, that's how it will start. IIRC, mail was the first driver for commercial flight.
This early stage will help develop better engines and airframes, capable of flying faster and higher for less. From a Mach 7 high-altitude platform, it's a lot easier to launch something into orbit. As soon as it gets cheaper than the "payload on a stick" approach we currently use, we will start using it.
One other possible way to LEO is - don't laugh - balloons. There was a company exploring this idea - launching a balloon to a high altitude and, from there, deploying a second, lighter balloon that would go up to the edge of the atmosphere carrying an ion-engine. It would then accelerate the balloon and payload to orbital velocities (since there would be little atmosphere, there is little friction). Since this is an unmanned launch system, there is not much of a hurry to get to LEO. Things get a lot more fuel-efficient if you don't need to be there in 3 minutes (which is where chemical rockets really shine).
"Is fixing all the messed up countries in the world OUR responsibility, though?"
Well... The US foreign policy has a long tradition of interfering deeply with ("messing up, of you wish") other countries to further its interests. I live in a country that had a US-backed military dictatorship for about 20 years because the US felt they should "save" us from a leftist president. Lucky me it wasn't nearly as big a bloodbath than the one installed in Chile, also with US support. The US also has a long tradition of ignoring messed up countries whose governments are favorable to their interests and only gets interested in "fixing" them when it fits their needs.
People should learn to fix what they break. If they don't like fixing, they shouldn't break stuff.
Slashdot Mirror
I would find it extremely cool to be sort of an undead person.
"Look: No pulse"
It's too bad I would be too ill to go to any interesting party...
I remember that, during my teens, I once woke up in a hospital after a party. The doctor, with an extremely serious face, explained to my half-conscious brain that I had too little blood in my alcoholstream and that the condition had been corrected.
I, for one, welcome our artificial-heart overlords.
Flamebait?!
They could easily subsidize the hardware required to run the system. Airlines would even demand they pay for the extra fuel the heavier computers would require ;-)
They could also go the thin client with beefy server route. Maybe not as good for a in flight entertainment system, but good enough for browsing and emailing.
But they will probably wait until people (read: their astroturf teams) start complaining OpenOffice is not Office and refuses to read their MSOOXML files before they announce their move that will "save" the airlines from the headaches of the previous system.
Never underestimate neither their weaseliness nor the depth of their pockets.
"linux is best in computers or embedded devices where you need high reliability and you want to be able to specify the exact amount of the functionality it should have."
While I would like to point out this is not about critical flight control systems (where I doubt any Linux would be certified as it costs a lot to be) and in-flight entertainment machines are OK to crash sometimes, the specific functionality is, probably, a win for Linux distros.
But, in the end, I suspect the real deal here is about price. The cheapest solution won. It would be hideously expensive to have Windows Vista PCs with Office 2007 on every seat of a jetliner.
+1 insightfuil/informative. Wish I had mod-points today...
http://apple2history.org/history/ah10.html
For multi-threading apps, instead of multiple cores (nothing except caches are shared between complete CPU cores), it makes a lot of sense to have an HT-like architecture (multiple context stores, shared elements) that reduces the time it takes to do a context-switch. It would also help a lot to have a context-aware cache system where a swapped-in context would not wake up having to read every instruction from main memory.
Since not all threads will be runnable at any given time, having more cores instead of bigger caches could hurt performance instead of helping it. A bunch of context stores and a couple independent CPU elements will not give you the same performance boost than a second core will, but they take up a lot less silicon and may spare you many trips to main memory.
Agreed.
The IIgs was too little, too late to save the Apple II line. It came out after the Mac, the Amiga 1000 and the Atari 520ST. At the time, it ran mostly 8-bit Apple II software (so, it was an overexpensive IIe that could run few software the older brother couldn't) and the full GS/OS only appeared much later. As things happened, it stood little to no chance of survival. In the end, the Apple IIe outlived it.
If you ask me, what doomed the II line was the failure of the IIx project. Had they finished it as a next-gen Apple II and launched it before the competition, they would have ensured the survival of the II line for a couple more years.
Not to say I don't want one. Really, I want all three in my interesting computer collection.
I'm in! ;-)
The II (6502) to Lisa/Mac (68K) was not a real transition. They had different lines of computers with different software. The Lisa or the Mac was never compatible with the II (without an expansion card). Even the Apple III, which used the same 6502, was barely compatible with the II.
The 68K to PPC transition was very smooth and much of the software base ran without a hitch. It worked so well only 8.6 was fully PPC and I bet a lot of it still ran under 68K emulation.
The PPC to x86 transition was interesting. Seemingly nothing changed, but the computers got a whole lot faster.
They would need a decent FPU/vector accelerator for this kind of application. I say it's very feasible, but, perhaps, not the best way to go.
OTOH, it would be much more clever to use MIPS or ARMs as general-purpose CPUs. On the same space, it should be feasible to cram a couple ARM11s, lots of cache and a couple DSP/SPU/GPU-like cores sitting around the main cores. It would no longer be as low-power as these x86s, but it would be a lot more interesting for desktop workstations. I find the Tilera 64-core chip very interesting in this regard. If they could build 32 CPU/FPU pairs instead of 64 FPU-less cores, it would be even more impressive. I would love to see a desktop made with them.
Best of all, they would be remarkably Windows-proof.
"If anyone ever tells you that switching ISA is prohibitive, go point them at Apple, because it worked out damn well for them"
Actually, they did it twice.
I still have my Color Classic around. It runs SSH just fine.
"There is no way you can implement a current x86 chip in the same number of transistors as a RISC chip like ARM or MIPS"
;-)
There... I corrected that for you.
the original x86 had less transistors than, probably, my wristwatch.
And you can implement a MIPS or ARM core with the same space you would a x86. It will just have 10x more cache
big and *very* evil.
Six to ten billion years ago? I don't think the Universe was hospitable to life at that time for a civilization to evolve and make a mess this huge.
"You don't have to be a phd to figure that out"
Not being properly educated is a sure way to thing something you don't quite understand is crap.
Big Bang is still a very good theory, but it's a theory. People will adjust it to match the observed facts and, when it becomes increasingly hard to adjust it, people will prefer other better theories that may fit the facts with less tweaking.
And no. "God and angels" is by no means a decent explanation.
Having to be a little be more specific proves my point (as you doubtlessly intended).
I like in Brazil.
Names like these happen when the guy who comes up with the name is a lot smarter than the guy who approves it.
"Apollo program got started by replacing the warheads with men in tin can"
I think that would be Mercury. All Apollo vehicles were ridiculously large for ICBMs.
"Or take shuttle. It was designed on paper, and the very first hardware iteration was declared operational configuration. Thats just nuts."
That happens when politics and engineering get too close. The shuttle has so many different mission objectives it could not (at this time with present technology) be a good solution.
As for this evolutionary approach, you need economic incentive. Passenger flights evolved because there were passengers willing to pay, airlines willing to buy and operate planes and, thus, companies willing to build them. There is no such thing for passenger spaceflight.
There _is_ for unmanned launchers and you see all kinds of evolution there. From the early launchers to Ariane V there is a lot between. There is little less variation than with airplanes mainly because the problem is a much more difficult one and there are fewer possible solutions with the same technologies. Right now, we can't make anti-gravity or warp-drive, so, we will have to stick to chemical rockets. And nuclear-thermal is a regulation nightmare. As much as I love the idea of cheap and widespread space-travel, I don't want people to have easy access to nuclear fuel.
If I would imagine a future with cheap access to space, I would start with commercial sub-orbital flight. If you figure out a way to reliably fly 60 passengers between China and California for a reasonable fare in a couple hours and without complex infrastructure at the ends, I think you would get filthy rich. Even if you could do so with cargo, DHL and FedEx would order everything you could build. If it does, that's how it will start. IIRC, mail was the first driver for commercial flight.
This early stage will help develop better engines and airframes, capable of flying faster and higher for less. From a Mach 7 high-altitude platform, it's a lot easier to launch something into orbit. As soon as it gets cheaper than the "payload on a stick" approach we currently use, we will start using it.
One other possible way to LEO is - don't laugh - balloons. There was a company exploring this idea - launching a balloon to a high altitude and, from there, deploying a second, lighter balloon that would go up to the edge of the atmosphere carrying an ion-engine. It would then accelerate the balloon and payload to orbital velocities (since there would be little atmosphere, there is little friction). Since this is an unmanned launch system, there is not much of a hurry to get to LEO. Things get a lot more fuel-efficient if you don't need to be there in 3 minutes (which is where chemical rockets really shine).
Wish I had modpoints...
"Is fixing all the messed up countries in the world OUR responsibility, though?"
Well... The US foreign policy has a long tradition of interfering deeply with ("messing up, of you wish") other countries to further its interests. I live in a country that had a US-backed military dictatorship for about 20 years because the US felt they should "save" us from a leftist president. Lucky me it wasn't nearly as big a bloodbath than the one installed in Chile, also with US support. The US also has a long tradition of ignoring messed up countries whose governments are favorable to their interests and only gets interested in "fixing" them when it fits their needs.
People should learn to fix what they break. If they don't like fixing, they shouldn't break stuff.
"As would giving anyone else that amount of money. And raising taxes to pay for it would remove an equal amount from the economy."
It's all about who you take from and to whom you give.
Just explain me how the fact that they are democrat and one of them speaks Arabic is relevant to the discussion?
You seem to regard "democrat" and "speaks fluent Arabic" as flaws of character, which they are not.