Why will the bond between the assembler and the product (or product part) be weakened, rather than one of the other bonds(say on the nanoassembler, or within the product) when strain is imposed?
Because it's designed that way. Specfically, you set up a system where a large, stable network of bonds on each side (a diamondoid latice of mostly carbon is the typical example) holds the atoms in question in place--but not quite the right place--to hold to the tool. There are many ways to acomplish this; you might be able to shade the location by doping the two latices, for example, or angle them with respect to each other. The latices provide deep energy wells to stablize the part and the assembler, much deeper than the energy required to break the bonds between them.
Here's a very loose analogy to help you picture how this would work:
Take two piles of objects and bind them both into bundles with lots of rubber bands. Now connect them to each other with a single rubber band. Now pull them apart...what happens?
Or even better, here's an example from real life (not, mind you, an analogy--this is an actual example of the exact effect under discussion):
Place a small bit of paper on the table. Press down on it with your finger, and then lift your finger off the table, taking the paper with it. Touch it to your tongue, and pull your finger away. Why did the bonds between your finger and the paper break, rather than the bonds within the paper or your finger? (Spit out the paper before answering).
It's a blurry line, but if I understand you correctly, you're saying you could (in the cell example) use nanotechnology to make the proteins themselves, but don't care where they float around in the cell itself, so long as they're in the general area. Like the mitochondria, the cell wall, etc etc.
Pretty much. The only exception is that if we needed to we could position the parts more exactly. Cells do this, by the way. They allow stuff to float at random only to the extent that exact positioning doesn't matter. Otherwise, molecules (or even medium large structures) are towed into position by active mechanisms, including ratchet-and-prawl devices that walk along micro-tubules.
but the chemistry at the site is generally most important and decided first, with the other factors also important but more easily adjustable.
Drop down a level in your thinking. From a physics perspective, "the chemistry at the site" disapears. It is, after all, just an aproximate way of describing the forces, charge distribution, etc.
The only reason it apears to be the most important factor is because we look at it that way, and have tools to deal with problems effectively when looked at from a chemical perspective. If we had a bit more omph in our computers (or better QM software) we could elliminate the chemical perspecive all together and just talk about the forces on various charged bodies. All we'd really need to take into acount is 1) charge, 2) mass, 3) the fact that electrons are fermions. Everything else a consequence.
hydrogen bonding, Van der Walls forces, dipole-dipole interactions, etc. and other 'weak' interactions simply cannot be ignored on this scale.
No one is saying they can be ignored. For that matter, Drexler doesn't ignore them (see Nanosystems for examples). You can't ignore friction when designing a car but that doesn't mean that the existence of friction makes cars impossible.
All atoms near other atoms interact... that's fundamental.
Right. But not all interactions are the same. Some are atractive, some repulsive, some strong and others weak, some highly dependent on angle or distance and others not as dependent. This is the stuff of which engineering is made. It's a known art. Build huge tables of interactions and all the data you can get on them, and let the designers pick and choose the combination that works best for them. Rinse, lather, and repeat.
Drexler thinks he can ignore all the other interactions to make an unreactive assembler work in a general case,
He most clearly does not. He thinks (and, as I've stated, I think he's correct) that these are engineering problems and not "basic science" problems. No one is saying they can be "ignored" anymore than they would say that things like drag can be ignored when designing an aircraft. But that doesn't mean (as Smalley seems to conclude) that the whole thing is impossible, anymore than the existence of drag makes aircraft impossible.
There are lots of potential unwanted side reactions, just as in any field of engineering, and there are lots of ways to deal with them, just as in any field of engineering. But you'll never get anywhere if all you do is throw up your hands and say "Look, complications! We're dombed to failure!"
The advances in chemistry alone needed to design bonds strong enough to hold parts yet still be weaker than the parts own internal bonds would make nanomachines nothing but simple toys. Instead of nanomachines we could just make a new liquid made of 80% nitrogen and 20% oxygen with bonds designed with just the right energy to be able to burn in a car engine, and then leave just N2 and O2 in the exhust.
Nuts. The way you make the bonds the strength you want to be is by choosing the atoms on either side and imposing strain (e.g. making them not line up quite right to make the bonds weak, or making them unable to move out of alignment to make the bond weak). You can also make them "switchable" by messing with charge distribution. This isn't something you could do with a liquid of pre-specified composition, but would be easy with a solid if you get to specify the structure and composition.
I sort of thought that one of the goals of nanotechnology is to produce something that can get every atom exactly where you want it.
A good working definition is the production of macroscopic quantities (and presumably useful assemblages) of complex components which are manufactured with every atom exactly where you want it.
The point is that you make sure you have your components built exactly the way you want them (so they do what you want) but you don't waste effort over specifying, trying to put every component "where you want it" to the same accuracy unless it serves some purpose. In fact, it would be bad design (IMHO) to try to do so, since this would lead to the sort of brittleness Smalley is so adamant about.
If we could accurately construct arbitrary "rational" diamondoid latice structures even 1/10 the scale of a typical cell we would have nanotech. (By rational I mean something along the lines of "excluding anything that is too unstable, or that can only be constructed by passing through one or more intermideate structures that are too unstable")
In other words, cells do nanotech, but they don't make nearly as many interesting things as I expect we will be able to.
Surely that sort of mechanism is more akin to biotechnology (and genetics) than nanotechnology.
No, biotech et al presuppose a mechanism. Nanotech is something different, in the same way that aviation is a field apart from ornithology.
Drexler never once explains exactly how he expects to hold onto these atoms & molecules without running into problems from interaction between the fragment to be manipulated and the mechanical arm's substrate.
Seems to me he's been explaining it for years. In short, the part would be bonded (either covaliently or ionicly) to the arm, with bonds that were designed to be strong enough to "hold" the part but much weaker (w. respect to forces on at least one axes) than the internal bonds on the part. When it was in the proper position with respect to the destination assembly, forces would be applied to create bonds between the part and it's new home. Now the whole thing is bonded together (including the arm). The arm would be freed by applying different forces (at right angles, say) to break the bonds between it and the part.
Dexler seems to live in a fantasy world where Heisenberg's Uncertainty Principle doesn't exist and the position and momentum of atoms can be precisely measured and controlled.
Nuts. He has addressed this point multiple times. Besides, in what follows you are talking about thermal, not quantum, positional uncertanty.
While it is theoretically possible to control a reaction if you can move atoms with perfect precision, you simply cannot do this, especially at any reasonable temperature, and exactly how useful is a process that can only happen in a vacuum at near absolute zero?
Cells do it at room temperature. So nature, at least, doesn't agree with you.
The energy involved in working at room temperature will make atoms jump around wildly and uncontrollably. (Creating pre-cooked food, my arse...)
Creating (frozen, say) pre-cooked food should be no different than creating raw food at the same temperature. And note that most of our food is created by creatures that do their work at room temperature, and (at the smallest scale) assemble every molecule with every atom in the right place. The fact that they get sloppy as the sub-assemblies get larger has more to do with the cost/benifit (to them) than to any technical limits.
*laugh* I can't imagine either of them talking the way we're putting words in their mouths, but since I started it I can't really complain.
The semantic issue isn't really "fingers" vs. "an aperatus of molecules", it's more "chemical" vs. "mechanical".
If at first you define "chemical" reacions to be ones that generally take place in a solution between molecules that meet at random angles and at random speeds, etc. then catalists and solvents are amazing (and important) things, while "mechanical" interactions seem to have nothing to do with it. When you finally get into the nitty gritty of what makes a catalist, you find that it is about mechanical things like aplying forces, constraining rotations, etc. When you understand the role of solvents you see that they are important because they do mechanical things like moving reactants around and possibly acting as co-factors for the catalists.
So then you have the option of redefining "chemistry" to be about reactions between molecules regardless of where and how they take place, and you can consider different ways of moving things around, aplying forces, constraining motion, etc. This is the leap Drexler has made but Smalley (I infer) has not.
The whole "fat/sticker" fingers thing is a gross over generalization of how atoms behave. Sure, some are very "sticky" (e.g. charged, or radicals, or just highly reactive due to unfilled orbitals), but it is also possible to design very non-reactive surfaces. Likewise, some structures are very "floppy" and therefore "fat" when you try to position them, but others are far stiffer (or actively track) and can be used for fine positioning.
The argument is about putting molecules together mechanically, as Drexler proposes. Drexler repeatedly refuses to address this point. I agree with Smalley that mechanical positioning, as Drexler advocates, is an inherently very limited method.
But it is literally the only method we've got. Even making soup requires you to (mechanically) bring the ingredients into proximity. Enzymes, etc. are just useful pre-existing means of mechanically applying various forces to speed up a reacion. Why stop there? I don't see it as Drexler refusing to address the point as much as his critics refusing to acknowledge his response.
It's most definitely not a straw man argument if it has the whole field of chemistry and biology behind it.
A "strawman" is an argument aimed at a misstatement of your opponent's possition. It doesn't matter who's "behind" it; at best, dragging in the issue makes it an argument by authority or ad hominim argument as well.
Cells can't make exact copies of themselves and I don't believe a group of cells working together would have any more luck in completing that task.
Exact? No. But functional? Sure. Cells do it all the time. In fact, birds do it. Bees do it. And from what I've been told, even educated fleas do it.
He's arguing that, without breakthroughs in chemistry beyond what we can currently imagine (ie some manner of "enzymes" not dependent on water or some other liquid), any molecular assembler that depends on the exact placement of individual atoms will suffer from tremendous error rates rendering the likelihood of creating a functional product infinitesimal.
Beyond what we can currently immagine? That's an argument from incredulity if I ever heard one. And a silly one, since we easily can imagine such things, and have examples (e.g. in gas phase) that come close to what is needed. As for the yields, the arguments get a little more complicated but it's really more a question of engineering than of basic science. There are all sorts of ways to deal with low yields, and armys of practitioners to apply them.
I think Smalley went a bit far in declaring molecular assemblers impossible
Yes. This is the essence of the disagreement. No one is saying we know how to build them now. But Smalley is (incorrectly, IMHO) jumping from this fact to the conculsion that there is no point in trying since the fact that we don't know how to build them now means that it is impossible to build them ever.
We can make macroscopic objects out of complex components already.
Strawman. The issue is complex components that have every atom where you want it (with perhaps some acceptance of a very low error rate).
Drexler's theory is that we can make an arbitrary object.
Strawman. The proposal is that we should be able to make arbitrary members of a huge class of useful objects, which isn't the same thing at all.
Does that mean we can use biochemical techniques to assemble macroscopic assemblies? No.
Strawman and argument from incredulity. First, this isn't Drexler's position, and second I see no reason why we couldn't, since trees do it on a regular basis.
We haven't the least clue how proteins form 3D shapes from their constituents. It's a great unsolved problem in biology and chemistry. The first one to solve it wins at least ONE Nobel prize.
Strawman. Ignoring the fact that we do know a great deal about how proteins fold, it doesn't matter since we don't need to understand the details anymore than the Wright brothers needed to understand how bird poop in order to build an airplane.
From current research, we know that we cannot self-assemble every molecule we can imagine. Some will self-assemble and some different types of assemblies are possible. But we're still a *long* way from being able to assemble an abitrary combination--which Drexler requires.
Argument from incredulity / ignorance (we don't know how to do it, therefore it can't be done) and Strawman.
And if you resort to what life can do, we're quite limited. Has life ever made a skyscraper?
Strawman.
>This is the same sort of plausible sounding arguments that have been used to "prove" (in my
>lifetime) that we will never detect planets around other stars,.... etc....
I see this argument all the time and it's totaly falacious. In my lifetime scientists have claimed that artificial inteligences will be so far superior to human inteligences that they will rule the world for us. When would this occur by? Well according to some 1960s AI pioneers, we'd be ruled by AIs by the 1980s.
Then you should read more carefully. What I said was that Smalley's arguments were invalid (specifically, that he raised strawmen and argued from incredulity, both of which are known types of faulty reasoning). I then listed a number of reasonably well known cases from the last few decades where this sort of reasoning had led to, what are in hindsight, wrong conclusions.
I never claimed or implied that the existence of a flawed argument against a premise is a evidence for the premise.
You imply that I hold this view, while I clearly state that I do not. I agree that the sort of argument that you describe is invalid, but that is not the sort of argument I made.
-- MarkusQ
P.S. If you are interested in an argument for Drexler's position that I find compelling, I can outline one:
I accept the matterialist premise that there is nothing "magical" about life; anything that a living creature can do can at least in principle be done by a machine of some type.
Living creatures build macroscopic objects out of complex components assembled to atomic precission.
Therefore, we should at least in principle be able to do so too.
You can't just say "well, this works with an iron atom in a hemoglobin molecule, so let's make a different carrier molecule with the same geometry, put it on a robot arm and use it to collect up nickel atoms, or whatever". Biology works because over billions of years a limited group of reactions has been found to work on a limited range of materials, in bulk and in carrier liquids. The notion that this means you can just build little tiny cranes and waggle atoms around does not follow.
Let's see how well that argument stands up in paraphrase:
You can't just say "well, this works with a bird in the sky, so let's make a different wing with the same shape, put it on a vehicle and use it to fly around, or whatever". Animal mobility works because over billions of years a limited group of structures has been found to work on a limited range of environments. The notion that this means you can just build airplanes does not follow.
No one is suggesting that we would blindly copy the geometry of some biological mollecule (without regard to it's charge distribution, orbital occupation, potentials for resonences, etc.), attach it to a robot arm, and expect it to do the job, any more than we would build an airplane by glueing birds wings on a school bus.
The whole biological-existence-proof line of arguments came up because some people (including Smalley, IIRC) claimed that building macroscopic objects out of components assembled with atomic precission was impossible in principle. Life forms are a clear concrete example of something that is build in exactly that way.
Now, saying that birds exist does not tell you how to build an airplane (though birds might be a good place to look for hints); all it does is shoot a big hole in the argument that flight is impossible.
If I had to bet, I'd say that Drexler was right. Smalley seems to rely on strawman arguments (they'd be restricted to water) and arguments from incredulity (the fat fingers schtick). This is the same sort of plausible sounding arguments that have been used to "prove" (in my lifetime) that we will never detect planets around other stars, that we will never be able to image individual atoms, that I will never have a hi-res colour display on my desk, that we will never be able to clone a mammal, etc., etc.
If you strip away the fancy words (and shamelessly simplify), this becomes much more obvious:
Drexler: We can build structures with atoms exactly where we want them, within reasonable limits.
Smalley:Your fingers are too big. Any robot you build will have fingers too big. It won't work.
Drexler: We wouldn't use "fingers," we'd use molecules designed for the purpose.
Smalley: I don't see how that could work.
Drexler: Living cells do it all the time.
Smalley: Ah, but they need water to do it. Your nano-things will only work in water.
And so forth...
Drexler may well be optimistic about the timeline, and may well be underestimating the difficulties, but I've yet to see an argument that it can't be done that holds up under critical examination.
HDD to DVD dubbing has been downgraded from 24x speed to 12x speed, CD-R playback will not be supported, DVD+RW playback will not be supported, MP3 file format will not be supported, TIFF and GIF file formats will not be supported, Sony's Cybershot movie format will not be supported. There will also be no antenna output, nor will it be compatible with the Playstation BB service.
Gosh, I hope it will still monitor my home network to make sure I'm not using raped-and-pillaged media files (I think that's what they call them) and making some poor pop-star starve. Otherwise the darned thing will be pretty much useless.
"Not so. For example, you can blackmail someone by threatening to tell the IRS that they didn't pay their taxes."
That's a false analogy, and I'm not convinced a court would find that to be blackmail at all. Just tell the judge you felt it was your duty as a citizen to report this heinous abuse, if it's true. If it's not true, it's slander of course.
You're missing the point. If you took the line you are proposing the judge would ask you why you offered not to tell if they paid you.
The essence of blackmail or extortion is the threat to do harm if (and only if) your demands aren't met.
SCO is saying that it owns something, that they are entitled to license terms, and *if* they choose to collect on those terms and you refuse to pay, then they will ask a court to settle the dispute. This isn't exactly the Giancana family putting the muscle on you.
The are claiming that they own something, but they refuse to say what it is. They are asking for money from people with whom they have had no prior dealings, and the only justification they give is the threat of a lawsuit, the basis of which they refuse to disclose.
It's only barratry if the case is without merit. Unless you're the judge in SCO v. IBM, I don't think you're qualified to make that conclusion just yet. Taking away SCO's right to petition the court to settle their dispute with IBM, is tantamount to abridging my own rights to due process of law. Until there is some ground to do so, and until there is some real evidence of wrongdoing by SCO, it's not right to accuse them.
Even if they win their case against IBM, this still doesn't give them anything against XYZ Corp. Even if they win against IBM, the hypothetical cases they are threatening random linux users with would have no merit. It is these, not the IBM case, that we are discussing here.
As far as it not being right to accue SCO until there is evidence, why shouldn't the same standard apply to them? On what basis do you defend their casting asperations far and wide without offering a shred of substantiation?
This might be closer to usury than it is to extortion.
Huh? I haven't seen any money lending at unreasonably high interest rates. This point makes no sense to me.
"Something bad might happen" must be something illegal.
Not so. For example, you can blackmail someone by threatening to tell the IRS that they didn't pay their taxes (which is not only legal but rewarded).
Barratry it may be, but that's not necessarily a problem under the current system.
Barraty is illegal. They may get away with it under the present system, but that is not the same thing as being legal.
If SCO were on the up-and-up they would demonstrate that they own something before charging people to use it, they would bill people for the use of their property rather than threatening to bill them, they would sue people who didn't pay the bills rather than threatening to sue them, and so forth.
The indemnity offer is not extortion, it's not a protection racket,
How do you figure? The essence of an extortion / protection racket is the bully goes to the victim and says "pay me or something bad might happen to you" and they both know that the thing that "might happen" will be caused to happen by the bully if they don't pay.
The bad thing could be anything from burning down the building (which is illegal) to a community boycott (which would in and of itself be perfectly legal). It could even be something (such as reporting and subsequent arrest of the victim for some crime that they actually committed) that the bully is legaly required to do. The nature of the threat doesn't matter. It's still extortion and/or a protection racket.
(In the particular case where the threat is of a lawsuit it's called "barity", IIRC).
It would be legal for SCO to bill people for using their software, or accuse people of stealing their software, but they are not doing this. Not only have they failed to point to any specific bit of software as "theirs", they even acknowledge that the (r) they put on Unix(r) indicates that it is a registered trademark of The Open Group, and not of SCO.
Demanding people pay you for using someone else's software or else you'll smear them in the press and annoy them in the courts is a protection racket.
I sometimes wonder if OSS will ever make any significant dent in the desktop market. Unfortunately the OSS community doesn't seem to get what's needed to make it happen.
Yes, it's amazing they haven't done as well as you have. I heard that thing you wrote was up to 20% of the desktops, with a bullet. Congradulations, and thanks again for taking time out of your busy day to coach the rest of us.
But seriously, the problem isn't that OSS people don't understand what the mass market wants, it's that we don't care. There are already lots of big companies bending over backwards to serve the needs of Joe Average. But we are too small (or perhaps just too fragmented) a market to be worth their while, so we wind up taking responsibility for serving our own needs ourselves. The ire you see directed at Microsoft, etc. is when they try to prevent this. It's the same reaction you'd get from (say) They Might Be Giants or Weird AL if you tried to make them shut up and listen to Brittany like everybody else.
I don't think the prior poster was being all that rude, especially by internet standards. But in the hopes of peace, goodwill, and understanding for all, I'll try to explain their point in (what I hope) will not seem to be a rude way.
Derivativation is something that courts have to decide.
Not really. They may have to decide in a questionable case, but having using a seperate name space instead of extending their schema as the poster suggested should be far from questionable. The whole point would be that you are going to reasonable lengths to avoid doing what they have said you can't do.
Imagine if Steven King were to proclaim that his next book could not be used as the basis for any "derived" works, and so Amazon decided that whenever anyone ordered a copy they would ship it in a seperate box from any other books that the customer ordered, invoice it seperately, etc. On what basis would you expect a court would object to this?
Are you willing to use namespaces in some code, deliver that code to a customer, then have Microsoft sue your customer for a patent violation?
Well, the risk wouldn't be all that great since: 1) it's a licence issue, not a patent issue, 2) the customer would not be held accountable since they hadn't done anything, 3) in the unlikely event that they attempted to hold you accountable for some claimed violation, you would have a strong case that you had made a good faith effort to abide by the terms.
Unless you have some sort of prescience, you shouldn't be taking that risk.
What standard of rude/not rude determination are you using here? At the very least, you might have tacked on an "IMHO", or stuck an "IANAL" somewhere.
--D00d, your friend is teh ROOT! Is she single??;-)
Last I heard she was "complex". It may be possible to cast her to single and then use the normal single operators on her, but everyone I've seen try this has seg faulted with a nasty access/privlege level violation message within a few seconds. Trying sleep() or u_sleep() is especially not recomended.
You can go a long way with this sort of thing...some might say too far.
When I was out to dinner with a group of friends a few years back one of them (who was a sysadmin for a medium sized company) interrrupted herself in mid sentence, pulled out her pager, looked at it, laughed, pulled out her PDA, made a few cryptic graffitesque gestures, said "Ha!", and pulled out her cell phone.
The conversation went something like this:
"Wally, why are you using Sara's computer?"
"Then you should have used one of the other machines in the sales department. You know the graphics people have asked the sales people not to use their machines without asking."
"That's right. And while I've got you on the line, do you happen to recall the company policy about browsing to inapropriate web sites?"
"Well, you know you shouldn't open attachments like that. Especially when you're using someone else's computer."
"Allright then. One last thing Wally: don't fuck with me."
Then she hung up and continued with what she'd been saying.
Slashdot Mirror
Why will the bond between the assembler and the product (or product part) be weakened, rather than one of the other bonds(say on the nanoassembler, or within the product) when strain is imposed?
Because it's designed that way. Specfically, you set up a system where a large, stable network of bonds on each side (a diamondoid latice of mostly carbon is the typical example) holds the atoms in question in place--but not quite the right place--to hold to the tool. There are many ways to acomplish this; you might be able to shade the location by doping the two latices, for example, or angle them with respect to each other. The latices provide deep energy wells to stablize the part and the assembler, much deeper than the energy required to break the bonds between them.
Here's a very loose analogy to help you picture how this would work:
Take two piles of objects and bind them both into bundles with lots of rubber bands. Now connect them to each other with a single rubber band. Now pull them apart...what happens?
Or even better, here's an example from real life (not, mind you, an analogy--this is an actual example of the exact effect under discussion):
Place a small bit of paper on the table. Press down on it with your finger, and then lift your finger off the table, taking the paper with it. Touch it to your tongue, and pull your finger away. Why did the bonds between your finger and the paper break, rather than the bonds within the paper or your finger? (Spit out the paper before answering).
-- MarkusQ
It's a blurry line, but if I understand you correctly, you're saying you could (in the cell example) use nanotechnology to make the proteins themselves, but don't care where they float around in the cell itself, so long as they're in the general area. Like the mitochondria, the cell wall, etc etc.
Pretty much. The only exception is that if we needed to we could position the parts more exactly. Cells do this, by the way. They allow stuff to float at random only to the extent that exact positioning doesn't matter. Otherwise, molecules (or even medium large structures) are towed into position by active mechanisms, including ratchet-and-prawl devices that walk along micro-tubules.
-- MarkusQ
but the chemistry at the site is generally most important and decided first, with the other factors also important but more easily adjustable.
Drop down a level in your thinking. From a physics perspective, "the chemistry at the site" disapears. It is, after all, just an aproximate way of describing the forces, charge distribution, etc.
The only reason it apears to be the most important factor is because we look at it that way, and have tools to deal with problems effectively when looked at from a chemical perspective. If we had a bit more omph in our computers (or better QM software) we could elliminate the chemical perspecive all together and just talk about the forces on various charged bodies. All we'd really need to take into acount is 1) charge, 2) mass, 3) the fact that electrons are fermions. Everything else a consequence.
hydrogen bonding, Van der Walls forces, dipole-dipole interactions, etc. and other 'weak' interactions simply cannot be ignored on this scale.
No one is saying they can be ignored. For that matter, Drexler doesn't ignore them (see Nanosystems for examples). You can't ignore friction when designing a car but that doesn't mean that the existence of friction makes cars impossible.
All atoms near other atoms interact... that's fundamental.
Right. But not all interactions are the same. Some are atractive, some repulsive, some strong and others weak, some highly dependent on angle or distance and others not as dependent. This is the stuff of which engineering is made. It's a known art. Build huge tables of interactions and all the data you can get on them, and let the designers pick and choose the combination that works best for them. Rinse, lather, and repeat.
-- MarkusQ
Drexler thinks he can ignore all the other interactions to make an unreactive assembler work in a general case,
He most clearly does not. He thinks (and, as I've stated, I think he's correct) that these are engineering problems and not "basic science" problems. No one is saying they can be "ignored" anymore than they would say that things like drag can be ignored when designing an aircraft. But that doesn't mean (as Smalley seems to conclude) that the whole thing is impossible, anymore than the existence of drag makes aircraft impossible.
There are lots of potential unwanted side reactions, just as in any field of engineering, and there are lots of ways to deal with them, just as in any field of engineering. But you'll never get anywhere if all you do is throw up your hands and say "Look, complications! We're dombed to failure!"
-- MarkusQ
The advances in chemistry alone needed to design bonds strong enough to hold parts yet still be weaker than the parts own internal bonds would make nanomachines nothing but simple toys. Instead of nanomachines we could just make a new liquid made of 80% nitrogen and 20% oxygen with bonds designed with just the right energy to be able to burn in a car engine, and then leave just N2 and O2 in the exhust.
Nuts. The way you make the bonds the strength you want to be is by choosing the atoms on either side and imposing strain (e.g. making them not line up quite right to make the bonds weak, or making them unable to move out of alignment to make the bond weak). You can also make them "switchable" by messing with charge distribution. This isn't something you could do with a liquid of pre-specified composition, but would be easy with a solid if you get to specify the structure and composition.
-- MarkusQ
I sort of thought that one of the goals of nanotechnology is to produce something that can get every atom exactly where you want it.
A good working definition is the production of macroscopic quantities (and presumably useful assemblages) of complex components which are manufactured with every atom exactly where you want it.
The point is that you make sure you have your components built exactly the way you want them (so they do what you want) but you don't waste effort over specifying, trying to put every component "where you want it" to the same accuracy unless it serves some purpose. In fact, it would be bad design (IMHO) to try to do so, since this would lead to the sort of brittleness Smalley is so adamant about.
If we could accurately construct arbitrary "rational" diamondoid latice structures even 1/10 the scale of a typical cell we would have nanotech. (By rational I mean something along the lines of "excluding anything that is too unstable, or that can only be constructed by passing through one or more intermideate structures that are too unstable")
In other words, cells do nanotech, but they don't make nearly as many interesting things as I expect we will be able to.
Surely that sort of mechanism is more akin to biotechnology (and genetics) than nanotechnology.
No, biotech et al presuppose a mechanism. Nanotech is something different, in the same way that aviation is a field apart from ornithology.
-- MarkusQ
Drexler never once explains exactly how he expects to hold onto these atoms & molecules without running into problems from interaction between the fragment to be manipulated and the mechanical arm's substrate.
Seems to me he's been explaining it for years. In short, the part would be bonded (either covaliently or ionicly) to the arm, with bonds that were designed to be strong enough to "hold" the part but much weaker (w. respect to forces on at least one axes) than the internal bonds on the part. When it was in the proper position with respect to the destination assembly, forces would be applied to create bonds between the part and it's new home. Now the whole thing is bonded together (including the arm). The arm would be freed by applying different forces (at right angles, say) to break the bonds between it and the part.
Dexler seems to live in a fantasy world where Heisenberg's Uncertainty Principle doesn't exist and the position and momentum of atoms can be precisely measured and controlled.
Nuts. He has addressed this point multiple times. Besides, in what follows you are talking about thermal, not quantum, positional uncertanty.
While it is theoretically possible to control a reaction if you can move atoms with perfect precision, you simply cannot do this, especially at any reasonable temperature, and exactly how useful is a process that can only happen in a vacuum at near absolute zero?
Cells do it at room temperature. So nature, at least, doesn't agree with you.
The energy involved in working at room temperature will make atoms jump around wildly and uncontrollably. (Creating pre-cooked food, my arse...)
Creating (frozen, say) pre-cooked food should be no different than creating raw food at the same temperature. And note that most of our food is created by creatures that do their work at room temperature, and (at the smallest scale) assemble every molecule with every atom in the right place. The fact that they get sloppy as the sub-assemblies get larger has more to do with the cost/benifit (to them) than to any technical limits.
-- MarkusQ
*laugh* I can't imagine either of them talking the way we're putting words in their mouths, but since I started it I can't really complain.
The semantic issue isn't really "fingers" vs. "an aperatus of molecules", it's more "chemical" vs. "mechanical".
If at first you define "chemical" reacions to be ones that generally take place in a solution between molecules that meet at random angles and at random speeds, etc. then catalists and solvents are amazing (and important) things, while "mechanical" interactions seem to have nothing to do with it. When you finally get into the nitty gritty of what makes a catalist, you find that it is about mechanical things like aplying forces, constraining rotations, etc. When you understand the role of solvents you see that they are important because they do mechanical things like moving reactants around and possibly acting as co-factors for the catalists.
So then you have the option of redefining "chemistry" to be about reactions between molecules regardless of where and how they take place, and you can consider different ways of moving things around, aplying forces, constraining motion, etc. This is the leap Drexler has made but Smalley (I infer) has not.
The whole "fat/sticker" fingers thing is a gross over generalization of how atoms behave. Sure, some are very "sticky" (e.g. charged, or radicals, or just highly reactive due to unfilled orbitals), but it is also possible to design very non-reactive surfaces. Likewise, some structures are very "floppy" and therefore "fat" when you try to position them, but others are far stiffer (or actively track) and can be used for fine positioning.
The argument is about putting molecules together mechanically, as Drexler proposes. Drexler repeatedly refuses to address this point. I agree with Smalley that mechanical positioning, as Drexler advocates, is an inherently very limited method.
But it is literally the only method we've got. Even making soup requires you to (mechanically) bring the ingredients into proximity. Enzymes, etc. are just useful pre-existing means of mechanically applying various forces to speed up a reacion. Why stop there? I don't see it as Drexler refusing to address the point as much as his critics refusing to acknowledge his response.
-- MarkusQ -- MarkusQ
It's most definitely not a straw man argument if it has the whole field of chemistry and biology behind it.
A "strawman" is an argument aimed at a misstatement of your opponent's possition. It doesn't matter who's "behind" it; at best, dragging in the issue makes it an argument by authority or ad hominim argument as well.
-- MarkusQ
Drexler is making handwaving claims that this will be doable, without actualy coming up with a concrete explanation of how it could be done.
Hardly handwaving.
If you have specific objections, raise them. But don't call something handwaving just because you haven't read it.
-- MarkusQ
Cells can't make exact copies of themselves and I don't believe a group of cells working together would have any more luck in completing that task.
Exact? No. But functional? Sure. Cells do it all the time. In fact, birds do it. Bees do it. And from what I've been told, even educated fleas do it.
He's arguing that, without breakthroughs in chemistry beyond what we can currently imagine (ie some manner of "enzymes" not dependent on water or some other liquid), any molecular assembler that depends on the exact placement of individual atoms will suffer from tremendous error rates rendering the likelihood of creating a functional product infinitesimal.
Beyond what we can currently immagine? That's an argument from incredulity if I ever heard one. And a silly one, since we easily can imagine such things, and have examples (e.g. in gas phase) that come close to what is needed. As for the yields, the arguments get a little more complicated but it's really more a question of engineering than of basic science. There are all sorts of ways to deal with low yields, and armys of practitioners to apply them.
I think Smalley went a bit far in declaring molecular assemblers impossible
Yes. This is the essence of the disagreement. No one is saying we know how to build them now. But Smalley is (incorrectly, IMHO) jumping from this fact to the conculsion that there is no point in trying since the fact that we don't know how to build them now means that it is impossible to build them ever.
-- MarkusQ
Strawman. The issue is complex components that have every atom where you want it (with perhaps some acceptance of a very low error rate).
Strawman. The proposal is that we should be able to make arbitrary members of a huge class of useful objects, which isn't the same thing at all.
Strawman and argument from incredulity. First, this isn't Drexler's position, and second I see no reason why we couldn't, since trees do it on a regular basis.
Strawman. Ignoring the fact that we do know a great deal about how proteins fold, it doesn't matter since we don't need to understand the details anymore than the Wright brothers needed to understand how bird poop in order to build an airplane.
Argument from incredulity / ignorance (we don't know how to do it, therefore it can't be done) and Strawman.
-- MarkusQ
You wrote: Then you should read more carefully. What I said was that Smalley's arguments were invalid (specifically, that he raised strawmen and argued from incredulity, both of which are known types of faulty reasoning). I then listed a number of reasonably well known cases from the last few decades where this sort of reasoning had led to, what are in hindsight, wrong conclusions.
I never claimed or implied that the existence of a flawed argument against a premise is a evidence for the premise.
You imply that I hold this view, while I clearly state that I do not. I agree that the sort of argument that you describe is invalid, but that is not the sort of argument I made.
-- MarkusQ
P.S. If you are interested in an argument for Drexler's position that I find compelling, I can outline one:
You can't just say "well, this works with an iron atom in a hemoglobin molecule, so let's make a different carrier molecule with the same geometry, put it on a robot arm and use it to collect up nickel atoms, or whatever". Biology works because over billions of years a limited group of reactions has been found to work on a limited range of materials, in bulk and in carrier liquids. The notion that this means you can just build little tiny cranes and waggle atoms around does not follow.
Let's see how well that argument stands up in paraphrase:
No one is suggesting that we would blindly copy the geometry of some biological mollecule (without regard to it's charge distribution, orbital occupation, potentials for resonences, etc.), attach it to a robot arm, and expect it to do the job, any more than we would build an airplane by glueing birds wings on a school bus.The whole biological-existence-proof line of arguments came up because some people (including Smalley, IIRC) claimed that building macroscopic objects out of components assembled with atomic precission was impossible in principle. Life forms are a clear concrete example of something that is build in exactly that way.
Now, saying that birds exist does not tell you how to build an airplane (though birds might be a good place to look for hints); all it does is shoot a big hole in the argument that flight is impossible.
-- MarkusQ
If I had to bet, I'd say that Drexler was right. Smalley seems to rely on strawman arguments (they'd be restricted to water) and arguments from incredulity (the fat fingers schtick). This is the same sort of plausible sounding arguments that have been used to "prove" (in my lifetime) that we will never detect planets around other stars, that we will never be able to image individual atoms, that I will never have a hi-res colour display on my desk, that we will never be able to clone a mammal, etc., etc.
If you strip away the fancy words (and shamelessly simplify), this becomes much more obvious:
Drexler may well be optimistic about the timeline, and may well be underestimating the difficulties, but I've yet to see an argument that it can't be done that holds up under critical examination.-- MarkusQ
HDD to DVD dubbing has been downgraded from 24x speed to 12x speed, CD-R playback will not be supported, DVD+RW playback will not be supported, MP3 file format will not be supported, TIFF and GIF file formats will not be supported, Sony's Cybershot movie format will not be supported. There will also be no antenna output, nor will it be compatible with the Playstation BB service.
Gosh, I hope it will still monitor my home network to make sure I'm not using raped-and-pillaged media files (I think that's what they call them) and making some poor pop-star starve. Otherwise the darned thing will be pretty much useless.
-- MarkusQ
You're missing the point. If you took the line you are proposing the judge would ask you why you offered not to tell if they paid you.
The essence of blackmail or extortion is the threat to do harm if (and only if) your demands aren't met.
The are claiming that they own something, but they refuse to say what it is. They are asking for money from people with whom they have had no prior dealings, and the only justification they give is the threat of a lawsuit, the basis of which they refuse to disclose. Even if they win their case against IBM, this still doesn't give them anything against XYZ Corp. Even if they win against IBM, the hypothetical cases they are threatening random linux users with would have no merit. It is these, not the IBM case, that we are discussing here.As far as it not being right to accue SCO until there is evidence, why shouldn't the same standard apply to them? On what basis do you defend their casting asperations far and wide without offering a shred of substantiation?
Huh? I haven't seen any money lending at unreasonably high interest rates. This point makes no sense to me.-- MarkusQ
- "Something bad might happen" must be something illegal.
- Barratry it may be, but that's not necessarily a problem under the current system.
If SCO were on the up-and-up they would demonstrate that they own something before charging people to use it, they would bill people for the use of their property rather than threatening to bill them, they would sue people who didn't pay the bills rather than threatening to sue them, and so forth.Not so. For example, you can blackmail someone by threatening to tell the IRS that they didn't pay their taxes (which is not only legal but rewarded).
Barraty is illegal. They may get away with it under the present system, but that is not the same thing as being legal.
What they are doing is extortion.
-- MarkusQ
The indemnity offer is not extortion, it's not a protection racket,
How do you figure? The essence of an extortion / protection racket is the bully goes to the victim and says "pay me or something bad might happen to you" and they both know that the thing that "might happen" will be caused to happen by the bully if they don't pay.
The bad thing could be anything from burning down the building (which is illegal) to a community boycott (which would in and of itself be perfectly legal). It could even be something (such as reporting and subsequent arrest of the victim for some crime that they actually committed) that the bully is legaly required to do. The nature of the threat doesn't matter. It's still extortion and/or a protection racket.
(In the particular case where the threat is of a lawsuit it's called "barity", IIRC).
It would be legal for SCO to bill people for using their software, or accuse people of stealing their software, but they are not doing this. Not only have they failed to point to any specific bit of software as "theirs", they even acknowledge that the (r) they put on Unix(r) indicates that it is a registered trademark of The Open Group, and not of SCO.
Demanding people pay you for using someone else's software or else you'll smear them in the press and annoy them in the courts is a protection racket.
-- MarkusQ
You can stop shouting. I'm pretty sure Diebold's heard you already.
-- MarkusQ
I sometimes wonder if OSS will ever make any significant dent in the desktop market. Unfortunately the OSS community doesn't seem to get what's needed to make it happen.
Yes, it's amazing they haven't done as well as you have. I heard that thing you wrote was up to 20% of the desktops, with a bullet. Congradulations, and thanks again for taking time out of your busy day to coach the rest of us.
But seriously, the problem isn't that OSS people don't understand what the mass market wants, it's that we don't care. There are already lots of big companies bending over backwards to serve the needs of Joe Average. But we are too small (or perhaps just too fragmented) a market to be worth their while, so we wind up taking responsibility for serving our own needs ourselves. The ire you see directed at Microsoft, etc. is when they try to prevent this. It's the same reaction you'd get from (say) They Might Be Giants or Weird AL if you tried to make them shut up and listen to Brittany like everybody else.
-- MarkusQ
Why don't you stop being rude?
I don't think the prior poster was being all that rude, especially by internet standards. But in the hopes of peace, goodwill, and understanding for all, I'll try to explain their point in (what I hope) will not seem to be a rude way.
Derivativation is something that courts have to decide.
Not really. They may have to decide in a questionable case, but having using a seperate name space instead of extending their schema as the poster suggested should be far from questionable. The whole point would be that you are going to reasonable lengths to avoid doing what they have said you can't do.
Imagine if Steven King were to proclaim that his next book could not be used as the basis for any "derived" works, and so Amazon decided that whenever anyone ordered a copy they would ship it in a seperate box from any other books that the customer ordered, invoice it seperately, etc. On what basis would you expect a court would object to this?
Are you willing to use namespaces in some code, deliver that code to a customer, then have Microsoft sue your customer for a patent violation?
Well, the risk wouldn't be all that great since: 1) it's a licence issue, not a patent issue, 2) the customer would not be held accountable since they hadn't done anything, 3) in the unlikely event that they attempted to hold you accountable for some claimed violation, you would have a strong case that you had made a good faith effort to abide by the terms.
Unless you have some sort of prescience, you shouldn't be taking that risk.
What standard of rude/not rude determination are you using here? At the very least, you might have tacked on an "IMHO", or stuck an "IANAL" somewhere.
-- MarkusQ
*laugh*
I'll send her a link to this thread then, and let her decide.
-- MarkusQ
--D00d, your friend is teh ROOT! Is she single??
Last I heard she was "complex". It may be possible to cast her to single and then use the normal single operators on her, but everyone I've seen try this has seg faulted with a nasty access/privlege level violation message within a few seconds. Trying sleep() or u_sleep() is especially not recomended.
-- MarkusQ
You can go a long way with this sort of thing...some might say too far.
When I was out to dinner with a group of friends a few years back one of them (who was a sysadmin for a medium sized company) interrrupted herself in mid sentence, pulled out her pager, looked at it, laughed, pulled out her PDA, made a few cryptic graffitesque gestures, said "Ha!", and pulled out her cell phone.
The conversation went something like this:
Then she hung up and continued with what she'd been saying.-- MarkusQ