This sounds like a good plan I wonder if I can make some money off my invention. It too has space applications:
We all know that in a vacuum you have particles and anti-particles being constantly created and destroyed. Many of these are charged so: Set up an intense electric field to pull all the positively charged particles to one end and negative particles to the other. If you get your charge strong enough you will pull the electrons away from the anti electrons before they can annihilate and Bob's your uncle you've got a cheap source of antimatter!
Now who do I speak to to get funding for research into my solution to the world's energy problems? What? No-one? It's the energy companies keeping new science down I tell you! Conspiracy!
And if they're not US citizens they're not really people and so killing them isn't as bad? I understand the legal point here WRT the constitution, but are you aware of how things like this sound to others?
To summarize your post. "I realise that a company has screwed you over. You are an idiot because knowing what you know now you should have gone to someone else." I'm sorry but no! The point is they have been amoral and done something at could screw over other people. The world needs to hear this and the company should be dragged through the coals because of this. Not that it will happen mind and yes really the only way a consumer can attack a large company is to take their business elsewhere, but that is only painfull to a company if lots of people do it.
These countries are in our back yard and sphere of influence and we would definitely stop attackers before they launched (see: cuban missile crisis).
I'm not sure that is a good example. Let's (for the sake of argument) suppose that relationships between Brazil and Argentina get worse. To the point where is limited military conflict between them. Both of them either develop their own or buy off the market some cruise missiles, actually let's assume they both get lots. Should the US intervene to stop them? What gives the US the right to interfere? These are sovereign nations and what right (other than might makes right) does one country have to interfere with others? But let's suppose the US does interfere - this very intervention would then be the event that kicks off that attack on the US with an overwhelming number of warheads.
I believe the point of the article wasn't to say that there is something to currently worry about,I think it was simply saying that the current situation is potentially flawed.
You CANNOT build something cheaply that will be able to cross the Atlantic or Pacific and still be aimed at a target when it gets here.
These days guidance technology is so cheap it's just not funny. Plenty of amateurs have demonstrated guidance systems on their model planes that cost no more than a few hundred dollars. Let's face it the cost of a guidance system is a GPS receiver, a small processor and a few servo motors. Now I believe you're right that the complexities of building a plane capable of crossing the Atlantic or Pacific is no mean feat. However I do not believe it is that much more complex than say building a modern car - and there are plenty of countries that are capable of doing that and doing it in huge numbers at low cost. Again I don't think any of those are currently annoyed at the US enough to do anything like this, but you can bet they have thought of it.
Don't forget as well the costs to build this are basically pi*d + turbines, and the storage capacity would be pi*r^2; so economies of scale rapidly kick in - it makes great financial sense to build this HUGE! And taking a large amount of farmland or living space out of commission when there is all that unused ocean there just seems plain daft in comparison. What "job" the ocean does it still can do with this, not the same as if you tried to build something analogous on land.
Nope NTRs have high ISP but (compared to Chemical rockets) low thrust. The whole point of the ffirst or zeroth stage is high thrust (the SRBs for example are very high thrust, but low ISP). Nuclear rockets are great for a second stage or later but they can often have a thrust to weight ration http://www.projectrho.com/public_html/rocket/enginelist.php#table for data.
Now an Orion drive on the other hand, yes that's a great way to get off planet nuclear - but unlikely to be used.
I don't think you're qualified to talk about this subject if you can't understand the massive difference between a "technology demonstrator" and "a (flight ready) engine". For example - it could require vast amounts of support hardware - it's not uncommon on test engines to have things like the turbo pumps and tanks and control hardware completely external to the engine and if you tried to fly with them they'd be too heavy or wouldn't package correctly or any other a a thousand little problems. The materials almost certainly aren't rated to flight standard - at the very least a lot of testing needs doing. The test engine probably hasn't been designed to cope with all the failure modes a real engine would have to. There will be lots of work on things like the gas flow pasterns in the injectors, under real acceleration this is a major problem - likewise oscillations and similar effects. Packaging - I mentioned it earlier but it's a big thing, getting stuff where you want it for test purposes is very different to where you want it for centre of mass balancing vs where you want it for thermal management which is different for where you want it for safety reasons vs where you want it for aerodynamic purposes etc. I could go on but I suspect you're thinking like a software engineer not a mechanical engineer and even then not a very experienced software engineer if you think a technology demonstrator is comparable to a production part.
Look let's put this in slashdot terms: There's big difference between the "Hey guys look I can boot Linux on this mobile phone" and getting a linux phone to market. Or "Hey I can get an SQL database running on my desktop" and a backend solution suitable for a large internet website. If you prefer a slashdot car analogy what we have here is an engine block on a test stand, integrating this new one of a kind engine with a whole car and getting it to production is a far bigger job than getting the engine running on the test stand in the first place.
Can I get a default block on things I find offensive to children please? * Violence * Religion * Telephone Scams * Adverts to tacky products * politicians * The Daily Mail
Seriously why the focus on this one thing that some people think is bad for some other people? If you have a problem with receiving something, you fix it, the tools are out there and free! Don't make your problem my problem because of your ignorance and laziness.
HFT does add value. It adds liquidity. If you do not understand why liquidity is a good thing thing: Imagine a market with near zero liquidity. Shares are only traded when a buyer is matched to a seller. Let's suppose I need to sell some shares (to pay for a new roof on my house perhaps) and I go to the market and no-one is interested in buying any shares in business X that I am selling. It turns out no-one is interested because they either know nothing about that company or the people who happened to turn up that day were not interested in a boring oil business but wanted to invest their cash in one of these sexy new internet startup things they had heard of. Potentially I am screwed, unless someone buys them on a whim. In steps banker Bob who realises I'm about to sell these shares way below what they're worth and so snaps them up at a bargain price. Is banker Bob evil for snapping them up cheap? [If you think he is then recall that had he not stepped in I would potentially have not been able to sell them at all and would have to remain living in a house with a leaky roof because no-one would buy my shares.] Am I stupid for having my nest egg in the random fluctuations of the stock market?
If banker Bob in this situation is not evil, and if the stock market is at least in principle intended to be a place for risk aware investors then I don't see anything wrong with the above scenario. I'm going to assume you agree with me here and continue. What would make the situation above even better is if there were more people like banker Bob around who would bid amongst themselves to determine a fair price for the shares I am selling, so I would get a fair price. The same argument replies in reverse for if you want to buy shares for a company you want to invest in, added liquidity means more availability of shares for you to buy. It also stops you over paying for shares due to a short term availability problem.
Remember at it's heart HFT is doing nothing illegal or immoral. A HFT is equivalent to the spotty young upstart on the trading floor who is able to hear the orders better, think faster and jump higher to catch the other person's attention in the pit trading. I'm sure the old veterans also on the floor resent his speed eyesight and hearing but that doesn't make it wrong.
I think the problem is the original article was about an approach to add another route to parallelise an existing bit of arbitrary code. Fyngyr jumps in and says something along the lines of 'well if you just wrote with parallelism in mind in the first place when architecting the system you wouldn't have this problem.' A viewpoint I agree with and in fact I would extend to say that from my viewpoint most systems are easier to design to be parallel in the first place than to force single threaded onto them once you are trained by whatever methods to think in this way. Now is where I think the problem comes, most SW engineers are so used to working in single threaded worlds and are so trained to think in single threaded modes it's all they know. The education and languages are so geared to single threaded viewpoint that this is understandable. So we end up with Fyngyr saying 'just design it to be parallel in the first place' and yourself correctly pointing out that not all problems can be parallel. Then the reaction of most SW engineers is that this is a compiler problem and we're back to TFA. I'd rather say that there is middle ground. IME some parts of the problem are inherently parallel, some are inherently serial and forcing one to be something it isn't is plain silly. The fact that the language encourages you to iterate serially over an image one pixel at a time without being able to say this can be done in parallel is to me daft. The fact that when most people design a large system they get the single threaded version working first then try and add in parallelism is totally wrong. I could rant all day about this but I think you get my point. Which is that while not all large problems are trivially parallelisable most are at least to some limited (but still huge) extent, so the fact that it cannot be solved for the totally general case should not stop it from being a preliminary design consideration and the fact that it often isn't is I think a shameful declaration of the state of the software industry. Now the coarsely grained multi-threaded-ness I'm talking about here may become invalid when we've got 256K core processors, but we're a way off that yet.
As a side note to make my point, consider the times when your computer seems slow. What is it doing? Perhaps rendering a web page while your backup application is running in the background? Perhaps you also need your mail app to be checking the server at the same time as you've got music playing in the background. In any of the cases i can think of where my computer seems slow to complete a task it is a completely trivial problem and there is no excuse outside of poor system architecture. (this is of course ignoring the compute intense stuff that I might do like let's say transcoding video or rendering 3d, or playing computer games - oh wait, also trivially parallel.)
So even if they have defects, they still get shipped, despite the numerous automated and manual checks?
I'm not sure what you mean. No test can capture 100% of all possible faults especially if you include any of the faults that mean things work to begin with but then fail later on (e.g. a weak solder joint will work to begin with, but the increased current density will tend to exacerbate the weak point making it fail terminally). In any chip itself there will be hard to find points of failure (a metal contact problem causing the pipeline to not flow control properly for example). Or they could be a weak driver on a memory cell caused by an implantation fail that means that under hot temperature conditions the memory write doesn't always successfully occur. What if it is a shielding problem on one of your clocks so that one multiplier experiences cross-talk sufficient to corrupt data in a cold chip only when the PSU is working with components at the limit of their tolerance? You can find many of these faults but how much is it worth adding to the cost of the system to catch that last 1% or last 0.00001% of problems?
Production test is a hard probabilistic field where many of the problems are none obvious. At the end of the day all you can do is have the best test methodology available with careful monitoring of defect rates backed up by a solid returns policy.
If you believe reality is consistent, time does not travel in 2 directions in 1 universe.
Actually you raise an interesting point. If there are no CPT violations then time could travel in both directions and we would be none the wiser. Imagine the last few seconds I just experienced suddenly ran backwards - without CPT violations I would be back at the same initial starting conditions and then time goes forwards again and I'd have exactly the same outcome. Time could move forwards and backwards totally randomly and we would never be any wiser. With CPT violations of course that could not happen.
What any of this means for the randomness of quantum mechanics I have no idea - especially with the disproval of the hidden variables theory.
I think you misunderstand. Light only ever travels at the speed of light. However when in a medium (such as water or glass) the interactions of light with the medium (by absorption/re-emission) cause the effective group velocity of the light to be slower. One photon could pass through the block of glass and not interact with any atoms at all and therefore travel at the speed of light. However it is vastly more probable it will hit a silicon or oxygen atom and excite an electron which will then re-emit the photon a short time later.
Hopefully one day someone will produce a decent GPU that's free down to the Verilog level
You raise an interesting point, but how would you make money from it? AFAICT current open source software companies make their money from consultancy rather than the actual product. Also things like opencores never really got much momentum behind them thanks to the twitchy legal departments.
I think the bigger problem though is you'd need the tool chain to go with it, so perhaps the first step would be an open source verilog simulator and debugger. There's no hardware equivalent of gcc. I don't know this would happen any time soon because you need engineers who are equally good at hardware and software. Massively parallel software being notoriously difficult to write!
By your own logic, I can just rearrange your sentences around and say the students could just take an after-school chemistry course outside of school.
You missed my point. I'm saying that they should learn a bit of everything. If you're good at something, great if you're not then the worst thing is you get a bad grade. The alternative is that you get universities saying we want someone with x number of y grades, and people game the system to take x subjects which are very similar. This defeats the point. Why does the university care that you get top marks in physics and maths and chemistry and technology and further maths if 90% of those courses involve the same skills? It's better for both the student and universoity/employer if the schools actually test the whole of the student.
A lot of students come to high school already knowing what they want to do and where they want to go in life. Why should high school be an impediment to that? Again, as you say, there is *plenty* of time to figure out how to be "balanced" after school. So how is it any different when we look at it that way?
I can only speak for myself here but I agree. I went to college knowing what I wanted to do and went to Uni to do it and now 16+ years later I'm still doing it. But you know what, I'm glad I hated the guts out of my english teacher and her pointless compositions on pigs coming out of the swamp to kill people on the council estate. I'm glad I sucked at music so bad that I for a start learned to fail and had deal with it. I'm glad that I can talk to friends at uni studying music about music theory because some of them were quite hot. I'm also glad that I can enjoy Guitar Hero and just enjoy music rather than have it be something that I'm expected to be good at because everyone else who is in that class is. IMO the geeky courses attracted those with lack of social skills and if that's all you saw you would become insular. This is not a good thing; please try and convince me otherwise, but I see no evidence.
If the employer of the child/young adult/etc doesn't care about the child's knowledge then the employer is free to ignore their chemistry grade.
If the child has an interest in public speaking he is free to join a club. If the child is interested in art above what is taught in school then he is free to do it in his own time.
School shouldn't be easy, if it is not hard then the school is doing its job wrong. Future employers have a right to know that their mathematical whiz that they are planning to employ is rubbish at Music and English lit. Maybe it won't matter for the job, or maybe it will show someone who has problems they need ot be aware of because it will affect their ability to do the job. If the kid is showing an aptitude for engineering he should still be learning about Literature and Biology otherwise he'd end up a misbalanced individual (can I use the example of Sheldon?) Likewise I have no talent for music but I am glad I was forced to do it because if I hadn't then I probably wouldn't have met some of my more "interesting" friends. I hated foreign languages and was useless at them, but I still end up using the knowledge every time I go on holiday.
You have the rest of your life to become good at something. Schools should be teaching you the basicas of balanced society.
No the reason it's strapped on the side was so that the main engines (at the time of design viewed as the expensive bit) could be returned and re-used. The only way to do that was to have the engines as part of the orbiter. This meant any detachable fuel tank had to be either at the top of the stack or on the side. The wings were there to provide the cross range capability, the original shuttle design was a lifting body i.e. no wings.
Actually I'd argue that if you have the tech to build a Dyson sphere you've the technology to dismantle the star and make much more efficient use of the matter than merely burning it to helium. You'd want to dismantle the star and use it as a matter source for fuel and construction, the alternative is to leave it wastefully burning and then eventually exploding. What a waste when there's all that entropy there that can be used.
For those who don't get the reference it's a beat poem called Storm by Tim Minchin http://www.youtube.com/watch?v=HhGuXCuDb1U He's very much being derisive of that sort of viewpoint. Sorry to spoil the joke.
In cases like this I like to think of it as if it were a documentary where they filmed much much more than they showed but then slimmed it down for the audience. There could have been an hours delay but that would have been boring to watch. So the editor removed the waiting around, he also removed the dialogue where they talk about the delay to avoid confusing the audience. We see evidence of this all the time in TV where they will scramble the fighters and people who were on the bridge one moment are in combat suits and jumping into their fighter seconds later when in fact getting changed and getting to the launch bay could not have taken less than about 20 minutes. No Magic needed, just the editor removing the boring bits for us.
you can basically tack on as much as you want, what will suffer is your ability to accelerate unless you mount a bigger engine.
Which will then need more armour and reaction mass which will then need more engines which... All I need to do in any case then is get one of my 200 megatonne nukes within a few hundred kilometres of you and I've knocked out your ability to manoeuvre. Then I drop a 10 megatonne nuke within a kilometer of you and you're dead. Or I just launch a shotgun like blast of 10*1 megatonne nukes travelling towards you at 0.1c and you'll never be able to react in time.
Don't get me wrong there's a place for well armoured outposts that are fortresses, but they're in hollowed out asteroids that are well hidden until your carrier gets within a few hundred ks then I launch a nuke at you. Oh and no point trying to nuke me first, because you don't know which asteroids I built my bases in (especially if they are unmanned and have been for a couple of centuries). And even if you do randomly nuke asteroids the base in the centre is well enough armoured to survive a single nuke(and you can't afford to nuke every asteroid).
Mixed constraints. The big guns only fired fore and aft, they also had lots of little movable guns for local defence. This to me is quite realistic that particle/laser weapons could turn out to be hard to steer the beam itself so for small local defence you steer the whole weapon whereas for the big guns that probably have an accelerator the length of the entire ship you cannot steer the beam so you just steer the ship. If you can steer/reflect the beam then so (perhaps) can the enemy.
I remember the quote you're referring to. He was talking about explosions and engines in space. His logic was that if a crewed ship explodes then whatever atmosphere* is in that ship will be part of that explosion. The explosion will carry the gas/liquid throughout the local space where it will intersect with another ship. At the distances involved in the sequences in question ships were buffeted by the wake from that explosion. This will be felt in those ships as a sound (as the wave of exploding gas will hit the hull of the ship and transmit to the hull of that ship as sound). So there is sound in space when you have shockwaves carried by gas that was part of the exploded ship.** He also said that most of the engines in B5 were rocket based so if you in your ship crossed the wake of another ship you would have sound from their engines conducted to your hull. I would guess part of the safety protocols involved would mean the ship's computer would stop you from flying close enough to cause damage but the realities of space combat would mean you'd try and fly as close to the other ships as you could.
* Note that different species' ships exploding exploded with different colour. This he said was to show that they used different chemistry in their artmospheres. ** Likewise any atomic weapon we make for space combat will likely have some ballast in there to cause damage that isn't purely Gamma rays, so expect that if a nuke goes off near your ship then you would feel some of the blast either from the ballast hitting the side of your ship as a plasma, or if you're closer (but far enough away you aren't killed by the radiation) then the nuke will vaporise the outer layers of your armour and as that boils it will impart a thrust to your ship, felt by you as a sound.
Slashdot Mirror
This sounds like a good plan I wonder if I can make some money off my invention. It too has space applications:
We all know that in a vacuum you have particles and anti-particles being constantly created and destroyed. Many of these are charged so:
Set up an intense electric field to pull all the positively charged particles to one end and negative particles to the other. If you get your charge strong enough you will pull the electrons away from the anti electrons before they can annihilate and Bob's your uncle you've got a cheap source of antimatter!
Now who do I speak to to get funding for research into my solution to the world's energy problems?
What? No-one? It's the energy companies keeping new science down I tell you!
Conspiracy!
And if they're not US citizens they're not really people and so killing them isn't as bad?
I understand the legal point here WRT the constitution, but are you aware of how things like this sound to others?
To summarize your post.
"I realise that a company has screwed you over. You are an idiot because knowing what you know now you should have gone to someone else."
I'm sorry but no!
The point is they have been amoral and done something at could screw over other people. The world needs to hear this and the company should be dragged through the coals because of this.
Not that it will happen mind and yes really the only way a consumer can attack a large company is to take their business elsewhere, but that is only painfull to a company if lots of people do it.
These countries are in our back yard and sphere of influence and we would definitely stop attackers before they launched (see: cuban missile crisis).
I'm not sure that is a good example.
Let's (for the sake of argument) suppose that relationships between Brazil and Argentina get worse. To the point where is limited military conflict between them. Both of them either develop their own or buy off the market some cruise missiles, actually let's assume they both get lots.
Should the US intervene to stop them? What gives the US the right to interfere?
These are sovereign nations and what right (other than might makes right) does one country have to interfere with others?
But let's suppose the US does interfere - this very intervention would then be the event that kicks off that attack on the US with an overwhelming number of warheads.
I believe the point of the article wasn't to say that there is something to currently worry about,I think it was simply saying that the current situation is potentially flawed.
You CANNOT build something cheaply that will be able to cross the Atlantic or Pacific and still be aimed at a target when it gets here.
These days guidance technology is so cheap it's just not funny. Plenty of amateurs have demonstrated guidance systems on their model planes that cost no more than a few hundred dollars. Let's face it the cost of a guidance system is a GPS receiver, a small processor and a few servo motors.
Now I believe you're right that the complexities of building a plane capable of crossing the Atlantic or Pacific is no mean feat. However I do not believe it is that much more complex than say building a modern car - and there are plenty of countries that are capable of doing that and doing it in huge numbers at low cost.
Again I don't think any of those are currently annoyed at the US enough to do anything like this, but you can bet they have thought of it.
Don't forget as well the costs to build this are basically pi*d + turbines, and the storage capacity would be pi*r^2; so economies of scale rapidly kick in - it makes great financial sense to build this HUGE!
And taking a large amount of farmland or living space out of commission when there is all that unused ocean there just seems plain daft in comparison. What "job" the ocean does it still can do with this, not the same as if you tried to build something analogous on land.
Nope
NTRs have high ISP but (compared to Chemical rockets) low thrust. The whole point of the ffirst or zeroth stage is high thrust (the SRBs for example are very high thrust, but low ISP). Nuclear rockets are great for a second stage or later but they can often have a thrust to weight ration http://www.projectrho.com/public_html/rocket/enginelist.php#table for data.
Now an Orion drive on the other hand, yes that's a great way to get off planet nuclear - but unlikely to be used.
I don't think you're qualified to talk about this subject if you can't understand the massive difference between a "technology demonstrator" and "a (flight ready) engine".
For example - it could require vast amounts of support hardware - it's not uncommon on test engines to have things like the turbo pumps and tanks and control hardware completely external to the engine and if you tried to fly with them they'd be too heavy or wouldn't package correctly or any other a a thousand little problems.
The materials almost certainly aren't rated to flight standard - at the very least a lot of testing needs doing.
The test engine probably hasn't been designed to cope with all the failure modes a real engine would have to.
There will be lots of work on things like the gas flow pasterns in the injectors, under real acceleration this is a major problem - likewise oscillations and similar effects.
Packaging - I mentioned it earlier but it's a big thing, getting stuff where you want it for test purposes is very different to where you want it for centre of mass balancing vs where you want it for thermal management which is different for where you want it for safety reasons vs where you want it for aerodynamic purposes etc.
I could go on but I suspect you're thinking like a software engineer not a mechanical engineer and even then not a very experienced software engineer if you think a technology demonstrator is comparable to a production part.
Look let's put this in slashdot terms:
There's big difference between the "Hey guys look I can boot Linux on this mobile phone" and getting a linux phone to market. Or "Hey I can get an SQL database running on my desktop" and a backend solution suitable for a large internet website.
If you prefer a slashdot car analogy what we have here is an engine block on a test stand, integrating this new one of a kind engine with a whole car and getting it to production is a far bigger job than getting the engine running on the test stand in the first place.
Can I get a default block on things I find offensive to children please?
* Violence
* Religion
* Telephone Scams
* Adverts to tacky products
* politicians
* The Daily Mail
Seriously why the focus on this one thing that some people think is bad for some other people? If you have a problem with receiving something, you fix it, the tools are out there and free! Don't make your problem my problem because of your ignorance and laziness.
HFT does add value. It adds liquidity. If you do not understand why liquidity is a good thing thing:
Imagine a market with near zero liquidity. Shares are only traded when a buyer is matched to a seller. Let's suppose I need to sell some shares (to pay for a new roof on my house perhaps) and I go to the market and no-one is interested in buying any shares in business X that I am selling. It turns out no-one is interested because they either know nothing about that company or the people who happened to turn up that day were not interested in a boring oil business but wanted to invest their cash in one of these sexy new internet startup things they had heard of. Potentially I am screwed, unless someone buys them on a whim.
In steps banker Bob who realises I'm about to sell these shares way below what they're worth and so snaps them up at a bargain price.
Is banker Bob evil for snapping them up cheap? [If you think he is then recall that had he not stepped in I would potentially have not been able to sell them at all and would have to remain living in a house with a leaky roof because no-one would buy my shares.]
Am I stupid for having my nest egg in the random fluctuations of the stock market?
If banker Bob in this situation is not evil, and if the stock market is at least in principle intended to be a place for risk aware investors then I don't see anything wrong with the above scenario. I'm going to assume you agree with me here and continue.
What would make the situation above even better is if there were more people like banker Bob around who would bid amongst themselves to determine a fair price for the shares I am selling, so I would get a fair price. The same argument replies in reverse for if you want to buy shares for a company you want to invest in, added liquidity means more availability of shares for you to buy. It also stops you over paying for shares due to a short term availability problem.
Remember at it's heart HFT is doing nothing illegal or immoral. A HFT is equivalent to the spotty young upstart on the trading floor who is able to hear the orders better, think faster and jump higher to catch the other person's attention in the pit trading. I'm sure the old veterans also on the floor resent his speed eyesight and hearing but that doesn't make it wrong.
I think the problem is the original article was about an approach to add another route to parallelise an existing bit of arbitrary code.
Fyngyr jumps in and says something along the lines of 'well if you just wrote with parallelism in mind in the first place when architecting the system you wouldn't have this problem.' A viewpoint I agree with and in fact I would extend to say that from my viewpoint most systems are easier to design to be parallel in the first place than to force single threaded onto them once you are trained by whatever methods to think in this way.
Now is where I think the problem comes, most SW engineers are so used to working in single threaded worlds and are so trained to think in single threaded modes it's all they know. The education and languages are so geared to single threaded viewpoint that this is understandable.
So we end up with Fyngyr saying 'just design it to be parallel in the first place' and yourself correctly pointing out that not all problems can be parallel. Then the reaction of most SW engineers is that this is a compiler problem and we're back to TFA.
I'd rather say that there is middle ground. IME some parts of the problem are inherently parallel, some are inherently serial and forcing one to be something it isn't is plain silly. The fact that the language encourages you to iterate serially over an image one pixel at a time without being able to say this can be done in parallel is to me daft. The fact that when most people design a large system they get the single threaded version working first then try and add in parallelism is totally wrong. I could rant all day about this but I think you get my point. Which is that while not all large problems are trivially parallelisable most are at least to some limited (but still huge) extent, so the fact that it cannot be solved for the totally general case should not stop it from being a preliminary design consideration and the fact that it often isn't is I think a shameful declaration of the state of the software industry.
Now the coarsely grained multi-threaded-ness I'm talking about here may become invalid when we've got 256K core processors, but we're a way off that yet.
As a side note to make my point, consider the times when your computer seems slow. What is it doing? Perhaps rendering a web page while your backup application is running in the background? Perhaps you also need your mail app to be checking the server at the same time as you've got music playing in the background. In any of the cases i can think of where my computer seems slow to complete a task it is a completely trivial problem and there is no excuse outside of poor system architecture. (this is of course ignoring the compute intense stuff that I might do like let's say transcoding video or rendering 3d, or playing computer games - oh wait, also trivially parallel.)
Now you know how I've felt for the last year about the kindle fire.
So even if they have defects, they still get shipped, despite the numerous automated and manual checks?
I'm not sure what you mean. No test can capture 100% of all possible faults especially if you include any of the faults that mean things work to begin with but then fail later on (e.g. a weak solder joint will work to begin with, but the increased current density will tend to exacerbate the weak point making it fail terminally).
In any chip itself there will be hard to find points of failure (a metal contact problem causing the pipeline to not flow control properly for example). Or they could be a weak driver on a memory cell caused by an implantation fail that means that under hot temperature conditions the memory write doesn't always successfully occur. What if it is a shielding problem on one of your clocks so that one multiplier experiences cross-talk sufficient to corrupt data in a cold chip only when the PSU is working with components at the limit of their tolerance?
You can find many of these faults but how much is it worth adding to the cost of the system to catch that last 1% or last 0.00001% of problems?
Production test is a hard probabilistic field where many of the problems are none obvious. At the end of the day all you can do is have the best test methodology available with careful monitoring of defect rates backed up by a solid returns policy.
If you believe reality is consistent, time does not travel in 2 directions in 1 universe.
Actually you raise an interesting point. If there are no CPT violations then time could travel in both directions and we would be none the wiser.
Imagine the last few seconds I just experienced suddenly ran backwards - without CPT violations I would be back at the same initial starting conditions and then time goes forwards again and I'd have exactly the same outcome. Time could move forwards and backwards totally randomly and we would never be any wiser.
With CPT violations of course that could not happen.
What any of this means for the randomness of quantum mechanics I have no idea - especially with the disproval of the hidden variables theory.
I think you misunderstand. Light only ever travels at the speed of light.
However when in a medium (such as water or glass) the interactions of light with the medium (by absorption/re-emission) cause the effective group velocity of the light to be slower.
One photon could pass through the block of glass and not interact with any atoms at all and therefore travel at the speed of light. However it is vastly more probable it will hit a silicon or oxygen atom and excite an electron which will then re-emit the photon a short time later.
In unrelated news, prisoners are suing the prison board because orange "isn't a flattering colour and really clashes with my tattoos"
Hopefully one day someone will produce a decent GPU that's free down to the Verilog level
You raise an interesting point, but how would you make money from it? AFAICT current open source software companies make their money from consultancy rather than the actual product. Also things like opencores never really got much momentum behind them thanks to the twitchy legal departments.
I think the bigger problem though is you'd need the tool chain to go with it, so perhaps the first step would be an open source verilog simulator and debugger. There's no hardware equivalent of gcc.
I don't know this would happen any time soon because you need engineers who are equally good at hardware and software. Massively parallel software being notoriously difficult to write!
By your own logic, I can just rearrange your sentences around and say the students could just take an after-school chemistry course outside of school.
You missed my point.
I'm saying that they should learn a bit of everything. If you're good at something, great if you're not then the worst thing is you get a bad grade.
The alternative is that you get universities saying we want someone with x number of y grades, and people game the system to take x subjects which are very similar. This defeats the point. Why does the university care that you get top marks in physics and maths and chemistry and technology and further maths if 90% of those courses involve the same skills? It's better for both the student and universoity/employer if the schools actually test the whole of the student.
A lot of students come to high school already knowing what they want to do and where they want to go in life. Why should high school be an impediment to that? Again, as you say, there is *plenty* of time to figure out how to be "balanced" after school. So how is it any different when we look at it that way?
I can only speak for myself here but I agree. I went to college knowing what I wanted to do and went to Uni to do it and now 16+ years later I'm still doing it.
But you know what, I'm glad I hated the guts out of my english teacher and her pointless compositions on pigs coming out of the swamp to kill people on the council estate. I'm glad I sucked at music so bad that I for a start learned to fail and had deal with it. I'm glad that I can talk to friends at uni studying music about music theory because some of them were quite hot. I'm also glad that I can enjoy Guitar Hero and just enjoy music rather than have it be something that I'm expected to be good at because everyone else who is in that class is.
IMO the geeky courses attracted those with lack of social skills and if that's all you saw you would become insular. This is not a good thing; please try and convince me otherwise, but I see no evidence.
If the employer of the child/young adult/etc doesn't care about the child's knowledge then the employer is free to ignore their chemistry grade.
If the child has an interest in public speaking he is free to join a club. If the child is interested in art above what is taught in school then he is free to do it in his own time.
School shouldn't be easy, if it is not hard then the school is doing its job wrong. Future employers have a right to know that their mathematical whiz that they are planning to employ is rubbish at Music and English lit. Maybe it won't matter for the job, or maybe it will show someone who has problems they need ot be aware of because it will affect their ability to do the job.
If the kid is showing an aptitude for engineering he should still be learning about Literature and Biology otherwise he'd end up a misbalanced individual (can I use the example of Sheldon?)
Likewise I have no talent for music but I am glad I was forced to do it because if I hadn't then I probably wouldn't have met some of my more "interesting" friends. I hated foreign languages and was useless at them, but I still end up using the knowledge every time I go on holiday.
You have the rest of your life to become good at something. Schools should be teaching you the basicas of balanced society.
No the reason it's strapped on the side was so that the main engines (at the time of design viewed as the expensive bit) could be returned and re-used. The only way to do that was to have the engines as part of the orbiter. This meant any detachable fuel tank had to be either at the top of the stack or on the side.
The wings were there to provide the cross range capability, the original shuttle design was a lifting body i.e. no wings.
Actually I'd argue that if you have the tech to build a Dyson sphere you've the technology to dismantle the star and make much more efficient use of the matter than merely burning it to helium.
You'd want to dismantle the star and use it as a matter source for fuel and construction, the alternative is to leave it wastefully burning and then eventually exploding. What a waste when there's all that entropy there that can be used.
For those who don't get the reference it's a beat poem called Storm by Tim Minchin
http://www.youtube.com/watch?v=HhGuXCuDb1U
He's very much being derisive of that sort of viewpoint.
Sorry to spoil the joke.
.There's never more than a few seconds delay
In cases like this I like to think of it as if it were a documentary where they filmed much much more than they showed but then slimmed it down for the audience.
There could have been an hours delay but that would have been boring to watch. So the editor removed the waiting around, he also removed the dialogue where they talk about the delay to avoid confusing the audience.
We see evidence of this all the time in TV where they will scramble the fighters and people who were on the bridge one moment are in combat suits and jumping into their fighter seconds later when in fact getting changed and getting to the launch bay could not have taken less than about 20 minutes. No Magic needed, just the editor removing the boring bits for us.
you can basically tack on as much as you want, what will suffer is your ability to accelerate unless you mount a bigger engine.
Which will then need more armour and reaction mass which will then need more engines which...
All I need to do in any case then is get one of my 200 megatonne nukes within a few hundred kilometres of you and I've knocked out your ability to manoeuvre. Then I drop a 10 megatonne nuke within a kilometer of you and you're dead.
Or I just launch a shotgun like blast of 10*1 megatonne nukes travelling towards you at 0.1c and you'll never be able to react in time.
Don't get me wrong there's a place for well armoured outposts that are fortresses, but they're in hollowed out asteroids that are well hidden until your carrier gets within a few hundred ks then I launch a nuke at you. Oh and no point trying to nuke me first, because you don't know which asteroids I built my bases in (especially if they are unmanned and have been for a couple of centuries). And even if you do randomly nuke asteroids the base in the centre is well enough armoured to survive a single nuke(and you can't afford to nuke every asteroid).
Mixed constraints. The big guns only fired fore and aft, they also had lots of little movable guns for local defence.
This to me is quite realistic that particle/laser weapons could turn out to be hard to steer the beam itself so for small local defence you steer the whole weapon whereas for the big guns that probably have an accelerator the length of the entire ship you cannot steer the beam so you just steer the ship. If you can steer/reflect the beam then so (perhaps) can the enemy.
I remember the quote you're referring to. He was talking about explosions and engines in space.
His logic was that if a crewed ship explodes then whatever atmosphere* is in that ship will be part of that explosion. The explosion will carry the gas/liquid throughout the local space where it will intersect with another ship. At the distances involved in the sequences in question ships were buffeted by the wake from that explosion. This will be felt in those ships as a sound (as the wave of exploding gas will hit the hull of the ship and transmit to the hull of that ship as sound). So there is sound in space when you have shockwaves carried by gas that was part of the exploded ship.**
He also said that most of the engines in B5 were rocket based so if you in your ship crossed the wake of another ship you would have sound from their engines conducted to your hull. I would guess part of the safety protocols involved would mean the ship's computer would stop you from flying close enough to cause damage but the realities of space combat would mean you'd try and fly as close to the other ships as you could.
* Note that different species' ships exploding exploded with different colour. This he said was to show that they used different chemistry in their artmospheres.
** Likewise any atomic weapon we make for space combat will likely have some ballast in there to cause damage that isn't purely Gamma rays, so expect that if a nuke goes off near your ship then you would feel some of the blast either from the ballast hitting the side of your ship as a plasma, or if you're closer (but far enough away you aren't killed by the radiation) then the nuke will vaporise the outer layers of your armour and as that boils it will impart a thrust to your ship, felt by you as a sound.