Yeah true, but demolitions are not done with just one explosion 75% of the way up the building. There IS a lot of air inside, but there are also supports running all the way up.
Not those buildings - they were rather unusual in that they were mostly empty inside, with all support provided by the core and the outer walls, and long trusses between them supporting the floors. Very strong when everything's together, but with nothing providing secondary support if it fails.
[...] The only real interaction between neutrinos and the rest of the universe (or each other) is that they both create gravity and are subject to gravity.
Actually, neutrinos interact through the "weak" nuclear force, which is why they're usually described as "weakly interacting" - which is an unfortunately uninformative way to put it. This means that they have a specific, well defined, but rare interaction with ordinary matter that lets neutrino detectors find them.
What are called "forces" are actually types of fields generated by the presence of certain properties, such as mass or electric charge. The "weak" force is a field which allows certain types of particle transformations to take place within it, such as a neutron splitting into a proton and electron (beta decay).
It's not known if neutrinos have mass, but it's probable because mass allows them to change "flavour" and would account for the "missing neutrinos" from the sun, which appear to simply be changing to an undetectable flavour. In that sense they also interact gravitationally, but as you know, are too light to make up any important fraction of the dark matter, as well as being too light to detect gravitationally.
The Burroughs B5000 pretty much covered that. It was designed so that only "safe" compiler-emitted code could be run, because only a "safe" program could set the "safe" flag on an executable file (probably was a way around that for system development, but good enough for normal usage):
One thing that dismays me is that systems in the past had some amazing and wonderful technology beyond what most people these days would even imagine. Unfortunately, there was always a "simplification" trend: mainframes got refined and safer, but complex and expensive until someone figured out new and cheaper technology minicomputers could do 80% of the job for a fraction of the cost. Then minis encountered the same problems with the remaining 20% and had to have the same refinements added, becoming complex and expensive until newer and cheaper technology meant microcomputers could do 80% of the job for a fraction of the cost, and the cycle repeated. Except it got interrupted partway with PCs.
Unfortunately development could either do the same thing better, or do more things "well enough", and with PCs doing more things won out, leaving the people who came after to think that "good enough" is "best possible" (or it would have been changed by now, right?), and the pinnacle of computing accomplishments, whether hardware, OS, language, etc. is forgotten and becomes myth. Or is occasionally rediscovered as if it were new.
Anyway, the point is it's good to learn history, because you never know when the problem you have now will be one already solved decades ago.
Everything has to be compiled Position-independent, or pre-linked for a specific location
If it's a well designed processor, that's no problem. I remember programming 6809, position independent was as easy as position dependent (providing you don't move it once it starts running).
Virtual memory fragmentation as applications are loaded and unloaded
This isn't as easy on bare metal, as Mac OS 9 and earlier showed, but possible by relocating the position independent code or data as needed. Virtual machines (or adequate runtime environments) make it easy though, since you know which values are pointers, you can update them transparently (as happens during compacting garbage collection).
COW and paging get harder
You can still have paged mapped memory with a single address space. Maybe you lose a little efficiency, but still save on context switch overhead.
Where is the heap? Is there one? Or one per process?
Memory management would be different (discontinuous), but user level code should be using an intermediate library (like malloc/free) which hides all that anyway.
Not that I'm saying single address space is better, but a lot of the problems come mainly from using it with OS and environments designed for virtual address space - of course that's going to go badly. Legacy technology gets that way because it's more effort to change it than just adapt what came before, not because it's better.
It had to be subsonic to deploy the weapons. And arrows don't generally turn, so that was probably a good name for something that really only flies well straight.
There were maintenance issues as well, it wasn't designed with maintenance in mind. Some panels wouldn't fit right, and had to be re-riveted with new holes after being removed to get at the interior. Probably there would have been a workaround eventually, but its optimised performance certainly came at an operational cost, which was the secondary reason for canceling it (the development cost was the main reason, procurement cost third, all versus missile-based alternatives).
This happens elsewhere too. In the latest Israel-Hezballah conflict victims were
paraded
in a show for journalists.
Here's
another list of staged incidents, including one woman who's house was destroyed twice in the same day in two different locations. It even happens
in Canada
.
Before GUIs and the attempt to look like paper, this was pretty much universally accepted as the best colour combination. It's was used for WordPerfect, IDEs, BIOS configuration screens, custom applications, and others, and is the reason the Windows "Blue Screen of Death" is blue. Also partly why most VCR programming and setup screens are white on blue.
I'm amazed that knowledge so well known at the time has so completely disappeared that it's as if it never existed. GUIs took on other colour schemes for other reasons (what you see is what you get, which made the yellow (or white) on blue contrast badly), which isn't all bad, but certainly has lost a lot of utility.
As I understand it, the tradition started because it was considered sacreligious (or at least disrespectful) to vandalize or destroy a textual representation of the name, but since paper and other media have a finite life, almost any written representation would eventually wear out and be destroyed. Better to just not write it down in the first place and not worry about whether it's disrespectful when it gets worn out and illegible, scribbled over, or disposed of.
The usage model for heterogeneous processors is not that difficult. Graphics processors aren't far from this soft of thing already. Any specialised function can just be implemented as a driver - to install a new service that uses a specialised CPU you just install a the driver for it, and access it through standard I/O calls or an exposed API. Without the specialised processor, you'd emulate the slower function on the main CPU (or one of them) using a slower compatibility driver.
Compatibility, flexibility, ease of use, no problem.
Privacy legislation makes it illegal for the Government of Canada to maintain a unified database on people in Canada. A few years the Auditor General found that the Department of Human Resources accidentally built one, and it had to be dismantled (if you recall, HR also got into trouble for spending billions on employment grants and not keeping track of it, before the program was transferred to the Department of Industry, which basically has better databases for programs like that).
But this is being funded outside the British government, it's an EU project. With European investors bankrolling British technology companies, maybe something can get done.
None of these issues indicate that explicit instruction level paralellism (EPIC) is a bad idea.
No, but the fact that compile-time determination of instruction scheduling can't adapt to:
Execution profiles that differ at runtime from what the compiler assumed
Different CPU resources, such as more or fewer FPUs, load/store units, etc, than were foreseen when the instruction set was designed
...limits the flexibility of CPU design improvements. For example, maybe having two high speed specialty adders that can't subtract can greatly speed up certain code. Scheduling instructions at run-time lets the CPU do that, but if your instruction set assumed that all integer ALUs can add and subtract, you have to either include extra functions (maybe slowing them down to regular speed and killing the advantage), or give up on the idea.
CPU design has progressed steadily from static to dynamic optimisation. Complex memory-based indirect addressing modes have been replaced by simple loads and stores, complex multi-function instructions were broken down to simpler basic operations - in both cases, this allowed better optimisation simply because there is more flexibility. The problem with VLIW, Itanium included, has always been that it's a step back from dynamic to static decisions - EPIC just got rid of the no-ops by encoding the VLIWs a bit better.
Itanium does have a lot of other really nifty things about it, but those were not new. Its speed comes from the DSP world, and almost all of its features are just to compensate for the problems DSPs have trying to deal with non-DSP issues like interrupts and exceptions, caches (DSPs hate non-uniform memory), function calls, and so on. That ends up making it almost as fast as a much simpler CPU with a DSP coprocessor, for only a few times more cost.
There's no way to know for sure what the motivations behind anything are (except military stuff tends to get classified so you don't hear about it - if Iran has a supercomputer for nuclear research, why do you think you'd ever hear about it?). However, Iran is one of the top Islamic countries when it comes to science and promoting research. Iranians have the third highest number of peer-reviewed published research papers among Islamic countries, behind Egypt and Turkey, and the largest recent increase. So investment in science is obviously considered an important public policy in Iran.
Obviously part of this is to support the military. Iran has been trying to build a domestic arms industry, including helicopters and fighter jets - mostly copies of Western designs, but with more variation from the original as they get more skilled at it. But that still requires scientists, and having a scientifically literate population can only have a positive influence on the country.
Ahmadinejad never says "Israel", he always refers to "the Zionist regime" within Palestine. There are very interesting historical reasons for that. I'm often dismayed that few people are willing to learn even as much about the region's history as someone widely considered a kook. What does that say about most people's opinions?
If the physics that says that a black hole will evaporate harmlessly almost immediately is wrong, then the physics that says a black hole will form at all is also wrong. So what is the risk?
As for socialized medicine, When a Canadian finds out they have something serious they come to the USA to get it fixed. If they stay in Canada and wait for the socialized medicine there, they die of their ailment before their turn comes up.
I don't know what makes this informative, since none of it is true. Most Canadians simply couldn't afford to pay the enormous expense of serious medical treatment in the U.S any more than uninsured Americans can, could they?
The Canadian medical system works very well, relatively speaking, with two main problems, both related to funding. The first is the discovery twenty or thirty years ago that deficit financing of governments simply doesn't work, and having hit that wall, all levels of government had to cut back budgets severely, including those for health care (as well as make a few other boneheaded decisions leading to a shortage of doctors and nurses). The second is the tug-of-war between socialist and capitalist ideologies using the medical care system as a proxy for the battle.
Generally, the conservative governments tend to underfund the medical system so that they can point to the problems they've caused and say that privatisation will fix it. Or they avoid fixing the problems in the medical system for fear of being accused of moving towards private health care. The socialists just avoid fixing problems because ideologically that would imply it's not perfect the way it is.
But governments that let the problems get too bad get voted out, or they finally fix things to avoid losing the next election - the most conservative province has one of the better government funding records, and one of the longest serving governments. So overall, the Canadian health care system is very good, but with some ugly problems.
That being said, rich people get what they want where they can, including American health care. But that is not typical for the average Canadian.
You mean they're finally putting the Mach microkernel API emulation capabilities to use, after using it as nothing but a Unix implementation platform? Sounds like what Tannenbaum has been advocating for years (decades?) now.
Established technology can be replaced if a technology for another use becomes popular enough, and can be adapted to replace the existing technology, and has some advantage. For email, two candidate technologies are notification feeds like RSS or Atom, and social web sites like Myspace and Facebook.
Both have an advantage over email in that you can control who you receive messages from because the sender identity cannot be faked. In RSS, you poll to get updates, so you know with certainty who you are polling. Social web sites let you define your relationship to others, and require an identity that satisfies those running the sites.
Either one could be modified to function like email, and I expect they will in time. Once that happens, existing email will have little reason to exist, and will go the way of Usenet newsgroups - generally used for special purposes, to unpopular for spammers to bother with. Email spam will be dead.
See also Chavez; taking the farms from the white owners left a lot of land to work, and at gunpoint it gets worked quite poorly, lowering the amount of food for the populace.
This is not Chavez - he's quite pro-farmer. Maybe you're thinking of Zimbabwe. But you should check your facts.
America after World War Two was magnanimous; we had freed a billion people, almost completely for free (the Brits had a lend-lease thing going on) then we started pumping in millions for all the cities we'd just blown up
At the end of that war, the German POWs held by the U.S were re-classified as "Disarmed Enemy Combatants" (familiar?) so that the Geneva conventions no longer applied to them. This allowed them to be used as slave labour. In addition, prisoners appear to have been mistreated and intentionally starved, killing tens of thousands (France and the U.S.S.R did similar things to post-war prisoners that they held). It's been alleged that reports of these atrocities inspired the U.S Congress to implement the Marshall Plan to reconstruct Europe.
America has never said it wants to attack, change the government and own another nation;
That's only technically correct because of the "and", and only if "own" means direct territorial control. Many nations have been conquered and turned over to proxy leaders obedient to the U.S, from Panama (once to annex it from Colombia, again to arrest its leader) to Iran to Cuba (lost those to revolutions) to Iraq (and later Iraq again). And Chile, Nicaragua, Honduras, Philippines - the list is embarrassingly long.
Another case where you should check your facts.
The president of Iran for example has spoken many times of using a nuke to wipe Israel off the planet (in direct violation of UN law) so many times, we're pretty sure he means it.
Ahmadinajad has spoken of using a referendum to remove Israel from the map (combining it with the Palestinian territories as determined by all people there - kind of disingenuous given the population differences would skew a referendum badly against Israel/"the Zionists", but at least honest about it). I think this is another example of "big lie" persuasion - repeat something, even provably false, often enough, and people will believe it.
Another case where you should check your facts.
It's just so surreal, though; knowing the good we've done, the 40,000 men who died to clear France for example, the play-by-the-rules military that we have,
It's fairly good about rules, but in a war bad things happen. American soldiers have raped, then killed familes to cover it up, executed Iraqis and planted weapons on them, left guns, ammunition, and grenades lying on a street, staked out the place, and shot anyone who carries it away (whether for insurgent use, or to turn over to the police, didn't matter, they were shot). They've tortured and beaten to death a taxi driver who was accused of terrorism - and found innocent after the fact. Not to mention turning prisoners over to Iraqis who are known to torture - using the techniques learned under Saddam Hussein, but now with U.S blessing.
The war has gone on so long that rules are often becoming just suggestions, and it's getting worse.
Again, some facts might be useful. There are no absolutes, even for your favourite country.
Slashdot Mirror
Not those buildings - they were rather unusual in that they were mostly empty inside, with all support provided by the core and the outer walls, and long trusses between them supporting the floors. Very strong when everything's together, but with nothing providing secondary support if it fails.
[...] The only real interaction between neutrinos and the rest of the universe (or each other) is that they both create gravity and are subject to gravity.
Actually, neutrinos interact through the "weak" nuclear force, which is why they're usually described as "weakly interacting" - which is an unfortunately uninformative way to put it. This means that they have a specific, well defined, but rare interaction with ordinary matter that lets neutrino detectors find them.
What are called "forces" are actually types of fields generated by the presence of certain properties, such as mass or electric charge. The "weak" force is a field which allows certain types of particle transformations to take place within it, such as a neutron splitting into a proton and electron (beta decay).
It's not known if neutrinos have mass, but it's probable because mass allows them to change "flavour" and would account for the "missing neutrinos" from the sun, which appear to simply be changing to an undetectable flavour. In that sense they also interact gravitationally, but as you know, are too light to make up any important fraction of the dark matter, as well as being too light to detect gravitationally.
I think the (much, much delayed) Sun "Rock" SPARC-based CPU does this. Maybe the design delay indicates why it's not tried more often.
One thing that dismays me is that systems in the past had some amazing and wonderful technology beyond what most people these days would even imagine. Unfortunately, there was always a "simplification" trend: mainframes got refined and safer, but complex and expensive until someone figured out new and cheaper technology minicomputers could do 80% of the job for a fraction of the cost. Then minis encountered the same problems with the remaining 20% and had to have the same refinements added, becoming complex and expensive until newer and cheaper technology meant microcomputers could do 80% of the job for a fraction of the cost, and the cycle repeated. Except it got interrupted partway with PCs.
Unfortunately development could either do the same thing better, or do more things "well enough", and with PCs doing more things won out, leaving the people who came after to think that "good enough" is "best possible" (or it would have been changed by now, right?), and the pinnacle of computing accomplishments, whether hardware, OS, language, etc. is forgotten and becomes myth. Or is occasionally rediscovered as if it were new.
Anyway, the point is it's good to learn history, because you never know when the problem you have now will be one already solved decades ago.
Problems with SAS:
If it's a well designed processor, that's no problem. I remember programming 6809, position independent was as easy as position dependent (providing you don't move it once it starts running).
This isn't as easy on bare metal, as Mac OS 9 and earlier showed, but possible by relocating the position independent code or data as needed. Virtual machines (or adequate runtime environments) make it easy though, since you know which values are pointers, you can update them transparently (as happens during compacting garbage collection).
You can still have paged mapped memory with a single address space. Maybe you lose a little efficiency, but still save on context switch overhead.
Memory management would be different (discontinuous), but user level code should be using an intermediate library (like malloc/free) which hides all that anyway.
Not that I'm saying single address space is better, but a lot of the problems come mainly from using it with OS and environments designed for virtual address space - of course that's going to go badly. Legacy technology gets that way because it's more effort to change it than just adapt what came before, not because it's better.
I read an excellent short story on that subject here: I don't know, Timmy, being God is a big responsibility
FYI: http://www.tsa.gov/blog/2008/05/so-what-exactly-happens-to-all-of-that.html
There were maintenance issues as well, it wasn't designed with maintenance in mind. Some panels wouldn't fit right, and had to be re-riveted with new holes after being removed to get at the interior. Probably there would have been a workaround eventually, but its optimised performance certainly came at an operational cost, which was the secondary reason for canceling it (the development cost was the main reason, procurement cost third, all versus missile-based alternatives).
I agree with the MiG-25 comparison.
This happens elsewhere too. In the latest Israel-Hezballah conflict victims were paraded in a show for journalists. Here's another list of staged incidents, including one woman who's house was destroyed twice in the same day in two different locations. It even happens in Canada .
I'm amazed that knowledge so well known at the time has so completely disappeared that it's as if it never existed. GUIs took on other colour schemes for other reasons (what you see is what you get, which made the yellow (or white) on blue contrast badly), which isn't all bad, but certainly has lost a lot of utility.
As I understand it, the tradition started because it was considered sacreligious (or at least disrespectful) to vandalize or destroy a textual representation of the name, but since paper and other media have a finite life, almost any written representation would eventually wear out and be destroyed. Better to just not write it down in the first place and not worry about whether it's disrespectful when it gets worn out and illegible, scribbled over, or disposed of.
Compatibility, flexibility, ease of use, no problem.
http://www.fin.gc.ca/treaties/USA_e.html
Interestingly, the tax treaty rules supersede both Congress and Parliament tax laws, in areas where there's a conflict.
Privacy legislation makes it illegal for the Government of Canada to maintain a unified database on people in Canada. A few years the Auditor General found that the Department of Human Resources accidentally built one, and it had to be dismantled (if you recall, HR also got into trouble for spending billions on employment grants and not keeping track of it, before the program was transferred to the Department of Industry, which basically has better databases for programs like that).
But this is being funded outside the British government, it's an EU project. With European investors bankrolling British technology companies, maybe something can get done.
CPU design has progressed steadily from static to dynamic optimisation. Complex memory-based indirect addressing modes have been replaced by simple loads and stores, complex multi-function instructions were broken down to simpler basic operations - in both cases, this allowed better optimisation simply because there is more flexibility. The problem with VLIW, Itanium included, has always been that it's a step back from dynamic to static decisions - EPIC just got rid of the no-ops by encoding the VLIWs a bit better.
Itanium does have a lot of other really nifty things about it, but those were not new. Its speed comes from the DSP world, and almost all of its features are just to compensate for the problems DSPs have trying to deal with non-DSP issues like interrupts and exceptions, caches (DSPs hate non-uniform memory), function calls, and so on. That ends up making it almost as fast as a much simpler CPU with a DSP coprocessor, for only a few times more cost.
Obviously part of this is to support the military. Iran has been trying to build a domestic arms industry, including helicopters and fighter jets - mostly copies of Western designs, but with more variation from the original as they get more skilled at it. But that still requires scientists, and having a scientifically literate population can only have a positive influence on the country.
Ahmadinejad never says "Israel", he always refers to "the Zionist regime" within Palestine. There are very interesting historical reasons for that. I'm often dismayed that few people are willing to learn even as much about the region's history as someone widely considered a kook. What does that say about most people's opinions?
If the physics that says that a black hole will evaporate harmlessly almost immediately is wrong, then the physics that says a black hole will form at all is also wrong. So what is the risk?
The Canadian medical system works very well, relatively speaking, with two main problems, both related to funding. The first is the discovery twenty or thirty years ago that deficit financing of governments simply doesn't work, and having hit that wall, all levels of government had to cut back budgets severely, including those for health care (as well as make a few other boneheaded decisions leading to a shortage of doctors and nurses). The second is the tug-of-war between socialist and capitalist ideologies using the medical care system as a proxy for the battle.
Generally, the conservative governments tend to underfund the medical system so that they can point to the problems they've caused and say that privatisation will fix it. Or they avoid fixing the problems in the medical system for fear of being accused of moving towards private health care. The socialists just avoid fixing problems because ideologically that would imply it's not perfect the way it is.
But governments that let the problems get too bad get voted out, or they finally fix things to avoid losing the next election - the most conservative province has one of the better government funding records, and one of the longest serving governments. So overall, the Canadian health care system is very good, but with some ugly problems.
That being said, rich people get what they want where they can, including American health care. But that is not typical for the average Canadian.
You mean they're finally putting the Mach microkernel API emulation capabilities to use, after using it as nothing but a Unix implementation platform? Sounds like what Tannenbaum has been advocating for years (decades?) now.
Someone has a Markov chain utility, apparently...
Both have an advantage over email in that you can control who you receive messages from because the sender identity cannot be faked. In RSS, you poll to get updates, so you know with certainty who you are polling. Social web sites let you define your relationship to others, and require an identity that satisfies those running the sites.
Either one could be modified to function like email, and I expect they will in time. Once that happens, existing email will have little reason to exist, and will go the way of Usenet newsgroups - generally used for special purposes, to unpopular for spammers to bother with. Email spam will be dead.
A link: Were Nazis Tortured in World War II?
That's only technically correct because of the "and", and only if "own" means direct territorial control. Many nations have been conquered and turned over to proxy leaders obedient to the U.S, from Panama (once to annex it from Colombia, again to arrest its leader) to Iran to Cuba (lost those to revolutions) to Iraq (and later Iraq again). And Chile, Nicaragua, Honduras, Philippines - the list is embarrassingly long.Another case where you should check your facts.
Ahmadinajad has spoken of using a referendum to remove Israel from the map (combining it with the Palestinian territories as determined by all people there - kind of disingenuous given the population differences would skew a referendum badly against Israel/"the Zionists", but at least honest about it). I think this is another example of "big lie" persuasion - repeat something, even provably false, often enough, and people will believe it.Another case where you should check your facts.
It's fairly good about rules, but in a war bad things happen. American soldiers have raped, then killed familes to cover it up, executed Iraqis and planted weapons on them, left guns, ammunition, and grenades lying on a street, staked out the place, and shot anyone who carries it away (whether for insurgent use, or to turn over to the police, didn't matter, they were shot). They've tortured and beaten to death a taxi driver who was accused of terrorism - and found innocent after the fact. Not to mention turning prisoners over to Iraqis who are known to torture - using the techniques learned under Saddam Hussein, but now with U.S blessing.The war has gone on so long that rules are often becoming just suggestions, and it's getting worse.
Again, some facts might be useful. There are no absolutes, even for your favourite country.