>>My point is that the defaults of the language are almost uniformly dangerous.
>>This claim has no factual basis.
>What? No? As in none at all?
You said the defaults of the language are uniformly dangerous.
No I equivocated. I said almost. And I stand by it.
So, why didn't they use java or smalltalk or LISP for that embedded, persistent, realtime code ??? I think you know the answer--
Yes, I know the answer;-)
those languages make tradeoffs that make them more or less useless for embedded or realtime applications.
Bzzzzt. You lose. Actually we've redesigned it to be a mixture of Java and C, and ditched C++ entirely... The submillisecond hard realtime stuff is C, and Java is perfectly fine for soft realtime, and embedded.
I understand that LISP or Smalltalk have been successfully used in similar situations as well.
I said most of the time , even within one program designers rarely need maximum speed at every point. My point is that the defaults of the language are almost uniformly dangerous.
This claim has no factual basis.
What? No? As in none at all? Lack of a garbage collector? No array bounds checking? No constraints on pointer arithmetic? Automatic type casts usually with no compiler warnings?
C++ was not designed from scratch, it was constrained
I prefer the word 'damaged'
by C compatibility. For the most part, this is the reason. Basically, getting rid of "sharp bits" wasn't the primary goal of the language.
You mean it's an eyesore, but because we know why, that's ok then?
Look I've lived at the sharp end of the language more than most. I've worked on multi-million lines of embedded, persistent, realtime code much of it in C++, and without the benefit of the STL. Trust me when I say C++ has some major issues. I've also sometimes had to maintain less clueful peoples code... all I can say is: yuck.
There is an argument that drivers would drive much more safely if there was a sharp point sticking out of the steering wheel. Any accident and you would die. The drivers will be really attentive. However to ensure the safety, the drivers are going to have to slow down... In the software world this transfers to less code written.
If you have written a compiler (maybe even if you haven't) you will realize that writing a switch statement involves a jump table. Avoiding jumps really helps speed,
Whether a switch is slow or not is both compiler and processor dependent; although most compilers don't optimise this much, switches are usually slow. But I don't see what that has to do with the syntax of the language.
so while your idea of using a nobreak option instead is intriguing, I personally like the way C++ encourages you to think about your switch statements to code them more efficiently.
So you like the naked sharp spinning things because the pending death concentrates the mind? You're perverse if you don't mind me saying.
You can always just use if-then-else-ifs to avoid the fall through problems when you don't want to think about what you are doing.
Look at the sheer number of complaints of about gnome, kde, mozilla, and related projects being to slow and you will realize that there is never enough speed. And those are just UI programs.
That's more to do with algorithms than anything else. You can make almost anything faster with a better algorithms, and the two ings: caching and hashing. I bet you anything not one of those programs use the best algorithm.
Frankly, your number pulling of "99% of software engineers" not needing the speed of C++ is just ridiculous and arbitrary.
I said most of the time, even within one program designers rarely need maximum speed at every point. My point is that the defaults of the language are almost uniformly dangerous. 95+% of the time, switch statements should not fall through; what's with making falling through the normal behaviour? I should put a 'nobreak' in to stop it breaking, not the other way around. I mean sure, I almost always remember it... these days...;-)
Yes, C++ does not prevent you from making the kinds of mistakes that you refer to
No language can do that. C++ is the equivalent of a machine shop with no guards around any of the machinery. Sure, after you've been cut a few times you learn to move in a way that avoids the sharp bits; but here's an idea, why aren't there guards around them, that you can remove if you are doing something that needs them out of the way?
Languages which do restrict you in this area *cough*java*cough* also restrict your creativity and power to accomplish your task.
Oh yeah right. Lack of creativity and power.;-) You must be the worst software engineer in the world if the language is getting in your way like that, but I don't buy that, you sound like you almost know what you're doing. But what's that smell? It's the Bull.
I mean, writing a compiler that can compile it is one thing, but i'm not sure that any reasonably percentage of the population actually can use it.
In my opinion, and I've done a LOT of both C and C++; C++ is just a rotten language. I've nothing against a launguage that can run fast. I just don't think that in these days of >2Ghz processors that 99% of software engineers need that much speed most of the time; and forcing engineers to code in a particular style that enforces speed but impacts reliability is just crazy.I'm fed up with my software core dumping, memory leaking and inappropriately casting. I'm tired of switch statements that fall through by default. I really hate off by one errors blowing away the end of an array. And I really hate the way its so easy to write software that is susceptible to buffer overrun attacks.
Let's face it, C++ is slightly warmed over 1973 technology, and it really, really shows.
I don't agree with you about his capabilities; he's good. But even if I did, I don't agree with your conclusions!
Rocket science isn't rocket science anymore!;-)
You can pretty much troll over to the NASA websites and download 50+ years of research. This works, this doesn't, you got to avoid getting your lox pipes damp... It's all there.
It still does, not because the principles are too advanced but because the raw materials are hideously
expensive
No. They aren't. The raw materials are reasonably cheap, or they can be. The cost per kg of the fuel is miniscule, about $10/kg of payload. The fuel tank is fairly expensive- they tend to be composite affairs, but they can sometimes be made from aluminum alloy if you have two stages. Turbopumps are very complex, but XCOR use a piston pump, and other designs are out there, like the rotary rocket pump.
You can knock together a small liquid fuelled rocket for about a couple of thousand. It's difficult, but not impossible.
and because the margin for error is enormous.
I think that its pretty comparable to any other aerospace activity- and that's how the FAA regulators are looking at it right now in fact (approximately).
But I'm not holding out hope for a $200 ticket on a space shuttle anytime soon.
That's unlikely. The cost of a transatlantic Concorde flight (about $5000-10,000) is more likely; and in the next 15 years is just about possible; although $100,000 may be more likely.
Yeah, and its cut the cost so much that they've saved money hand over fist.
Oops gotta watch the sarcasm.
The Russians can price an entire rocket for less than the cost of the external tank. Partly it's because their engineers are astonishingly cheap. The other part is because their hardware has been designed to actually be cheap to produce.
Arguably NASA designs equipment to be more expensive. For example the SRBs are made in parts because otherwise they couldn't be made in certain districts...
At the end of the day, the only testing that really counts is whether it works when you use it for real.
Any other metric is at best a proxy for the real thing. The space shuttle is currently slightly better than 99% reliable. Before Challenger the accident estimates were anywhere upto a million to one. Afterwards, sane people dropped it to between 50:1 and 200:1 or so.
The Buran had a measured 100% reliability, but since it has only been flown once, the true reliability is unknown.
p.s. great link; a truly great vehicle; I like it much more than the Shuttle design.
Well then the environment around him has changed, moving his views further towards the extreme without consulting him.
No, no. He's always been like this;-)
I wasn't aware that communism had a monopoly on far-left dogmatism.
His views aren't exactly far-left though; or atleast it depends on what axis you are on. He emphasises freedoms- this is more typically a right-wing thing. True analysis of somebody's position really depends on analysis along multiple axes- the normal left-right wing axis is a huge oversimplification of course.
Re:Politicking oneself into obscurity
on
The Stallman Factor
·
· Score: 4, Insightful
and since the rise to power of Linux he's slipped pretty much totally into the lunatic fringe.
No. As far as I can tell (and I remember the world before Linux), his position hasn't changed much in the last 10+ years.
Stallman is just another extremist with a useful ideal but no practical value.
Not quite sure how something can be a useful ideal, but at the same time of no practical value.
If it's useful it has to be of practical value. Unless you mean politically, but I don't think that Stallman resembles that remark.
He did write the GPL and GNU emacs. Without the GPL Linux would have been dead meat- the GPL is protecting Linux from Microsoft. Normally they'd buy a competing company or drop their prices, or bring out a lookalike product. None of these options are possible with the GPL.
Of course the GPL flowed from his idealism.
Stallman is as far on the extreme left of the software world...
You're basically implying that Stallman is a communist.
The difference between Stallman and communism is the difference between sex and rape. Communism is enforced collaboration. Free software is optional; of course he'd prefer it if you used it, but he's not enforcing it. The free software ideal avoids you raping him, and him raping you.
On the other hand the software capitalism as espoused by Microsoft is more like paying a prostitute. It seems like a good idea at the time, but its always worse in the cold light of day... (in the case of Microsoft when the auditors are banging at the door, and you never really know where their software has been;-) )
Your proposed laser propulsion system would consume the electrical output of every nuclear reactor in the United States operating at full power for 6 hours.
Oh, right. So you're taking a quoted figure from somebody else. Where they refer to a completely different system. Which launches a completely different sized payload, with hugely bigger lasers. Then multiplied it by 5 for no sensible reason. And then claimed that I said this, that it applies to the system I outlined???
You're a total, total asshole, you know that? A real smeghead.
NOW is it starting to dawn on you how far off the mark you are/were?
Right. 25MWh is the entire North American output of energy for 6 hours. Yes, its all clear now, you are totally incapable of understanding what you are writing about.
What colour is the sky in your world? Do you ever visit earth? Are you on drugs? If not, they can't hurt you now, you are seriously demented. You have lost the plot, gone gah-gah, flipped.
As I pointed out, this can be corrected with adaptive optics the same way that you can collimate the beam of a few billion laser pointers. The only thing is it's easier to do with fewer lasers (to a point) than billions of them.
With multiple lasers the heating effects are smaller near to the ground, so it needs less correction, the semiconductor lasers are cheaper and probably more efficient. I also disagree with your characterisation of semiconductor lasers as 'laser pointers'. As you very well know laser pointers are in the 10mw range. These lasers would be nearer to a watt.
I would point out the obvious cooling and control difficulties
Yes. So? Having multiple disparate lasers means the power density is lower in fact, and if anything easier to cool.
150KW propells something that weighs the same as an empty coke can to the edge of the atmosphere, but that's straight up.
Let's assume that's true (it sounds plausible). That's a coke can which is really light. Air drag losses per kg are inversely proportional to vehicle mass for a given shape. Normally it's reckoned you need 7% of the energy to get to orbital altitude, and the other 93% to get to orbit. I can easily believe that 1Mw/Kg is enough to get you there for reasonable masses.
Again, the cost of the semi-laser is not the issue. Big lasers cost big money but don't need all the fancy adaptive optics that a multi-billion laser array would need.
You need 'fancy optics' (whatever that means exactly) whatever you do. I think hundreds of separate lasers are easier than a few big ones- they disturb the air less; but still optics to point at the target vehicle are going to be very necessary; of course!
Then there's power, which you'venot my system! woefully underestimated by (a) not taking a realistic flight path mission profile into account actually I understand there's a special trajectory required- it goes up a long way up to avoid losing the payload over the horizon/losing too much in atmospherics and (b) not considering a worthwhile payload mass in conjunction with launch vehicle mass.Actually the figures I was given were ~500kg vehicle mass, 140kg payload, ~100 MW, hydrogen fuel, ISP 600, with throwaway tanks. There's a heat exchanger in it, and apparently it uses water rather than hydrogen at low altitude to give it enough thrust to actually take off. There probably ARE lots of loose ends, but you've found exactly none of them so far. I betcha (although the presenter never mentioned it) that he used a software simulation including atmospherics and trajectory optimisation features, so those figures are gonna work. I wish I was working on it actually, it sounds really, really interesting.
From your bent I'd have to guess that you work with semi-lasers on CD-ROM/DVDNo. We sell nothing to consumers. Actually, some of our laser equipment blows the ends off connectors; try giving that to consumers. I work in long distance fiber optic telecomms, multiwavelength stuff; people on site are semiconductor laser designers; I have a degree in physics and electronics, and yes I have studied lasers, and orbital mechanics.
100Mw for 15 minutes. That's 400,000 kwh. 1 Kwh costs $0.01-0.05. That makes: $4000-$20,000 of electricity for ~100kg.
Doh! I've multiplied where I should be dividing!
That's actually $250-1250 for 100kg! Now that's more like it! Feel free to add an extra 0 if you want, if you think it needs 1GW. All it does is raise the initial costs; the per-launch costs are pretty much unaffected as far as I can see; apart from amortising the laser cost.
Oh, the humanity. Please, please PLEASE do a little research before embarking on a debate about a subject which you don't seem all that informed about. Semiconductor lasers are not in the same class of power production as what would be needed for a laser-based propulsion system. A semiconductor laser might be great for a laser pointer or a DVD-ROM reader, but pushing megawatts or gigawatts of power is NOT what they do well. You need something more akin to a chemical reaction laser to generate this kind of power. Do a lookup on the EXCIMER laser to see what you'd need in order to make this project a reality.
Really? No. That's what you need if you want one big laser beam. We don't want that. In fact the atmospheric effects would be very detrimental- the beam would scatter horribly.
This concept is thousands of semiconductors all pointing in parallel through a single telescope at the vehicle; and then having, say a thousand of these telescopes. (You can play with the numbers, but that's the concept).
If you do the maths, its gonna come comfortably under a billion.
Power costs MONEY, and with the type of power you'd need, it may well outrun the cost of the hardware. This cost would be repeated FOR EVERY LAUNCH.
Gasp! Hadn't thought of that! I feel truly chastened. No truly. Let's see:
100Mw for 15 minutes. That's 400,000 kwh. 1 Kwh costs $0.01-0.05. That makes: $4000-$20,000 of electricity for ~100kg. This compares favourably with ~$2500+/kg (Russian Proton) that it currently costs to go into space, and very favourably indeed with the Space Shuttle costs. Of course this is only part of the costs, but still.
hint: 100MW is off by at least an order of magnitude
The figure I have seen quoted is about 1MW/kg, but it depends a bit on the system you're looking at. Do you have a source for your figure? Still, even if we were out by an order of magnitude on the power (and I don't agree we are; in fact I wonder what you're smoking), then the price would still come in comfortably under $1000/kg depending on the source of power chosen, and the cost of the other components. That's cheaper than all current launchers, and as I tried to explain in other postings; the system scales.
You're coming across as someone who's doing a lot of talking and very little thinking.
Do you think so? Oh dear. How sad; still atleast I don't come across sounding all negative.
p.s. come back when you've investigated the cost per watt of semiconductor lasers power. ($10-50/W- dropping at ~30% per year).
p.p.s. I work with semiconductor lasers as part of my day job; I understand optics.
p.p.p.s. this wasn't my idea (I wish), it came from a professional rocket scientist... you know, somebody who actually designs them for a living, for money, for the US government. It is entirely possible there's some flaw; but if so I don't think you've spotted it.
You need more of them to get the same output power.
What duty cycle do you reckon is needed- 100:1 perhaps? Do you think that 100x the lasers might be a tad expensive? Ok, I suppose you can probably run the laser at higher peak power, so its not quite that bad, but you're probably still going to need an order of magnitude more lasers. It was expensive already but this is gonna cost...
Slashdot Mirror
>>This claim has no factual basis.
>What? No? As in none at all?
You said the defaults of the language are uniformly dangerous.
No I equivocated. I said almost. And I stand by it.
So, why didn't they use java or smalltalk or LISP for that embedded, persistent, realtime code ??? I think you know the answer--
Yes, I know the answer ;-)
those languages make tradeoffs that make them more or less useless for embedded or realtime applications.
Bzzzzt. You lose. Actually we've redesigned it to be a mixture of Java and C, and ditched C++ entirely... The submillisecond hard realtime stuff is C, and Java is perfectly fine for soft realtime, and embedded. I understand that LISP or Smalltalk have been successfully used in similar situations as well.
This claim has no factual basis.
What? No? As in none at all? Lack of a garbage collector? No array bounds checking? No constraints on pointer arithmetic? Automatic type casts usually with no compiler warnings?
C++ was not designed from scratch, it was constrained
I prefer the word 'damaged'
by C compatibility. For the most part, this is the reason. Basically, getting rid of "sharp bits" wasn't the primary goal of the language.
You mean it's an eyesore, but because we know why, that's ok then?
Look I've lived at the sharp end of the language more than most. I've worked on multi-million lines of embedded, persistent, realtime code much of it in C++, and without the benefit of the STL. Trust me when I say C++ has some major issues. I've also sometimes had to maintain less clueful peoples code... all I can say is: yuck.
There is an argument that drivers would drive much more safely if there was a sharp point sticking out of the steering wheel. Any accident and you would die. The drivers will be really attentive. However to ensure the safety, the drivers are going to have to slow down... In the software world this transfers to less code written.
Whether a switch is slow or not is both compiler and processor dependent; although most compilers don't optimise this much, switches are usually slow. But I don't see what that has to do with the syntax of the language.
so while your idea of using a nobreak option instead is intriguing, I personally like the way C++ encourages you to think about your switch statements to code them more efficiently.
So you like the naked sharp spinning things because the pending death concentrates the mind? You're perverse if you don't mind me saying.
You can always just use if-then-else-ifs to avoid the fall through problems when you don't want to think about what you are doing.
You could, but that's horrible.
That's more to do with algorithms than anything else. You can make almost anything faster with a better algorithms, and the two ings: caching and hashing. I bet you anything not one of those programs use the best algorithm.
I said most of the time , even within one program designers rarely need maximum speed at every point. My point is that the defaults of the language are almost uniformly dangerous. 95+% of the time, switch statements should not fall through; what's with making falling through the normal behaviour? I should put a 'nobreak' in to stop it breaking, not the other way around. I mean sure, I almost always remember it... these days... ;-)
Yes, C++ does not prevent you from making the kinds of mistakes that you refer to
No language can do that. C++ is the equivalent of a machine shop with no guards around any of the machinery. Sure, after you've been cut a few times you learn to move in a way that avoids the sharp bits; but here's an idea, why aren't there guards around them, that you can remove if you are doing something that needs them out of the way?
Languages which do restrict you in this area *cough*java*cough* also restrict your creativity and power to accomplish your task.
Oh yeah right. Lack of creativity and power. ;-) You must be the worst software engineer in the world if the language is getting in your way like that, but I don't buy that, you sound like you almost know what you're doing. But what's that smell? It's the Bull.
In my opinion, and I've done a LOT of both C and C++; C++ is just a rotten language. I've nothing against a launguage that can run fast. I just don't think that in these days of >2Ghz processors that 99% of software engineers need that much speed most of the time; and forcing engineers to code in a particular style that enforces speed but impacts reliability is just crazy.I'm fed up with my software core dumping, memory leaking and inappropriately casting. I'm tired of switch statements that fall through by default. I really hate off by one errors blowing away the end of an array. And I really hate the way its so easy to write software that is susceptible to buffer overrun attacks.
Let's face it, C++ is slightly warmed over 1973 technology, and it really, really shows.
Seriously; the Russians are launching for about 1/20th of NASA. And their engineers are about 1/10 of the cost, if they're even that cheap.
Rocket science isn't rocket science anymore! ;-)
You can pretty much troll over to the NASA websites and download 50+ years of research. This works, this doesn't, you got to avoid getting your lox pipes damp... It's all there.
No. They aren't. The raw materials are reasonably cheap, or they can be. The cost per kg of the fuel is miniscule, about $10/kg of payload. The fuel tank is fairly expensive- they tend to be composite affairs, but they can sometimes be made from aluminum alloy if you have two stages. Turbopumps are very complex, but XCOR use a piston pump, and other designs are out there, like the rotary rocket pump.
You can knock together a small liquid fuelled rocket for about a couple of thousand. It's difficult, but not impossible.
and because the margin for error is enormous.
I think that its pretty comparable to any other aerospace activity- and that's how the FAA regulators are looking at it right now in fact (approximately).
But I'm not holding out hope for a $200 ticket on a space shuttle anytime soon.
That's unlikely. The cost of a transatlantic Concorde flight (about $5000-10,000) is more likely; and in the next 15 years is just about possible; although $100,000 may be more likely.
Oops gotta watch the sarcasm.
The Russians can price an entire rocket for less than the cost of the external tank. Partly it's because their engineers are astonishingly cheap. The other part is because their hardware has been designed to actually be cheap to produce.
Arguably NASA designs equipment to be more expensive. For example the SRBs are made in parts because otherwise they couldn't be made in certain districts...
Any other metric is at best a proxy for the real thing. The space shuttle is currently slightly better than 99% reliable. Before Challenger the accident estimates were anywhere upto a million to one. Afterwards, sane people dropped it to between 50:1 and 200:1 or so.
The Buran had a measured 100% reliability, but since it has only been flown once, the true reliability is unknown.
p.s. great link; a truly great vehicle; I like it much more than the Shuttle design.
"Return to castle wolfenstein."
Anyway the other one was funnier. But not as accurate. Sorry John.
Wolfenstein 3D, that is ;-)
In fact experimentally, Buran is ONE HUNDRED PERCENT reliable!
p.s. not sure where you got the 12 years of testing from... quite arguably it only has a few days of testing.
No, no. He's always been like this ;-)
I wasn't aware that communism had a monopoly on far-left dogmatism.
His views aren't exactly far-left though; or atleast it depends on what axis you are on. He emphasises freedoms- this is more typically a right-wing thing. True analysis of somebody's position really depends on analysis along multiple axes- the normal left-right wing axis is a huge oversimplification of course.
No. As far as I can tell (and I remember the world before Linux), his position hasn't changed much in the last 10+ years.
Stallman is just another extremist with a useful ideal but no practical value.
Not quite sure how something can be a useful ideal, but at the same time of no practical value.
If it's useful it has to be of practical value. Unless you mean politically, but I don't think that Stallman resembles that remark. He did write the GPL and GNU emacs. Without the GPL Linux would have been dead meat- the GPL is protecting Linux from Microsoft. Normally they'd buy a competing company or drop their prices, or bring out a lookalike product. None of these options are possible with the GPL.
Of course the GPL flowed from his idealism.
Stallman is as far on the extreme left of the software world...
You're basically implying that Stallman is a communist.
The difference between Stallman and communism is the difference between sex and rape. Communism is enforced collaboration. Free software is optional; of course he'd prefer it if you used it, but he's not enforcing it. The free software ideal avoids you raping him, and him raping you.
On the other hand the software capitalism as espoused by Microsoft is more like paying a prostitute. It seems like a good idea at the time, but its always worse in the cold light of day... (in the case of Microsoft when the auditors are banging at the door, and you never really know where their software has been ;-) )
You've set a fine example on how to do that.
You obviously haven't learned anything else.
It would be good if you actually understood more than the person you are trying to 'teach'...
Oh, right. So you're taking a quoted figure from somebody else. Where they refer to a completely different system. Which launches a completely different sized payload, with hugely bigger lasers. Then multiplied it by 5 for no sensible reason. And then claimed that I said this, that it applies to the system I outlined???
You're a total, total asshole, you know that? A real smeghead.
NOW is it starting to dawn on you how far off the mark you are/were?
Right. 25MWh is the entire North American output of energy for 6 hours. Yes, its all clear now, you are totally incapable of understanding what you are writing about.
What colour is the sky in your world? Do you ever visit earth? Are you on drugs? If not, they can't hurt you now, you are seriously demented. You have lost the plot, gone gah-gah, flipped.
With multiple lasers the heating effects are smaller near to the ground, so it needs less correction, the semiconductor lasers are cheaper and probably more efficient. I also disagree with your characterisation of semiconductor lasers as 'laser pointers'. As you very well know laser pointers are in the 10mw range. These lasers would be nearer to a watt.
I would point out the obvious cooling and control difficulties
Yes. So? Having multiple disparate lasers means the power density is lower in fact, and if anything easier to cool.
150KW propells something that weighs the same as an empty coke can to the edge of the atmosphere, but that's straight up.
Let's assume that's true (it sounds plausible). That's a coke can which is really light. Air drag losses per kg are inversely proportional to vehicle mass for a given shape. Normally it's reckoned you need 7% of the energy to get to orbital altitude, and the other 93% to get to orbit. I can easily believe that 1Mw/Kg is enough to get you there for reasonable masses.
Again, the cost of the semi-laser is not the issue. Big lasers cost big money but don't need all the fancy adaptive optics that a multi-billion laser array would need.
You need 'fancy optics' (whatever that means exactly) whatever you do. I think hundreds of separate lasers are easier than a few big ones- they disturb the air less; but still optics to point at the target vehicle are going to be very necessary; of course!
Then there's power, which you've not my system! woefully underestimated by (a) not taking a realistic flight path mission profile into account actually I understand there's a special trajectory required- it goes up a long way up to avoid losing the payload over the horizon/losing too much in atmospherics and (b) not considering a worthwhile payload mass in conjunction with launch vehicle mass. Actually the figures I was given were ~500kg vehicle mass, 140kg payload, ~100 MW, hydrogen fuel, ISP 600, with throwaway tanks. There's a heat exchanger in it, and apparently it uses water rather than hydrogen at low altitude to give it enough thrust to actually take off. There probably ARE lots of loose ends, but you've found exactly none of them so far. I betcha (although the presenter never mentioned it) that he used a software simulation including atmospherics and trajectory optimisation features, so those figures are gonna work. I wish I was working on it actually, it sounds really, really interesting.
From your bent I'd have to guess that you work with semi-lasers on CD-ROM/DVD No. We sell nothing to consumers. Actually, some of our laser equipment blows the ends off connectors; try giving that to consumers. I work in long distance fiber optic telecomms, multiwavelength stuff; people on site are semiconductor laser designers; I have a degree in physics and electronics, and yes I have studied lasers, and orbital mechanics.
Doh! I've multiplied where I should be dividing!
That's actually $250-1250 for 100kg! Now that's more like it! Feel free to add an extra 0 if you want, if you think it needs 1GW. All it does is raise the initial costs; the per-launch costs are pretty much unaffected as far as I can see; apart from amortising the laser cost.
Where to begin the flaying ;-)
Oh, the humanity. Please, please PLEASE do a little research before embarking on a debate about a subject which you don't seem all that informed about. Semiconductor lasers are not in the same class of power production as what would be needed for a laser-based propulsion system. A semiconductor laser might be great for a laser pointer or a DVD-ROM reader, but pushing megawatts or gigawatts of power is NOT what they do well. You need something more akin to a chemical reaction laser to generate this kind of power. Do a lookup on the EXCIMER laser to see what you'd need in order to make this project a reality.
Really? No. That's what you need if you want one big laser beam. We don't want that. In fact the atmospheric effects would be very detrimental- the beam would scatter horribly.
This concept is thousands of semiconductors all pointing in parallel through a single telescope at the vehicle; and then having, say a thousand of these telescopes. (You can play with the numbers, but that's the concept).
If you do the maths, its gonna come comfortably under a billion.
Power costs MONEY, and with the type of power you'd need, it may well outrun the cost of the hardware. This cost would be repeated FOR EVERY LAUNCH.
Gasp! Hadn't thought of that! I feel truly chastened. No truly. Let's see:
100Mw for 15 minutes. That's 400,000 kwh. 1 Kwh costs $0.01-0.05. That makes: $4000-$20,000 of electricity for ~100kg. This compares favourably with ~$2500+/kg (Russian Proton) that it currently costs to go into space, and very favourably indeed with the Space Shuttle costs. Of course this is only part of the costs, but still.
hint: 100MW is off by at least an order of magnitude
The figure I have seen quoted is about 1MW/kg, but it depends a bit on the system you're looking at. Do you have a source for your figure? Still, even if we were out by an order of magnitude on the power (and I don't agree we are; in fact I wonder what you're smoking), then the price would still come in comfortably under $1000/kg depending on the source of power chosen, and the cost of the other components. That's cheaper than all current launchers, and as I tried to explain in other postings; the system scales.
You're coming across as someone who's doing a lot of talking and very little thinking.
Do you think so? Oh dear. How sad; still atleast I don't come across sounding all negative.
p.s. come back when you've investigated the cost per watt of semiconductor lasers power. ($10-50/W- dropping at ~30% per year).
p.p.s. I work with semiconductor lasers as part of my day job; I understand optics.
p.p.p.s. this wasn't my idea (I wish), it came from a professional rocket scientist... you know, somebody who actually designs them for a living, for money, for the US government. It is entirely possible there's some flaw; but if so I don't think you've spotted it.
They should use rocket engines... power to weight ratio is *much* better. Once up they can be shutdown and a prop could be used.
What duty cycle do you reckon is needed- 100:1 perhaps? Do you think that 100x the lasers might be a tad expensive? Ok, I suppose you can probably run the laser at higher peak power, so its not quite that bad, but you're probably still going to need an order of magnitude more lasers. It was expensive already but this is gonna cost...
This effect can be made arbitrarily small. The vehicle is moving very fast; so it can't heat at that end.
OTOH, the energy density at the ground can be made aribitrarily small by spacing the lasers, so I don't think this is more than an annoyance.