The real Aspie... lacks the capacity to know he doesn't fit in. That's what makes him an Aspie.
Wide of the mark. What can identify someone as an Aspie is that he lacks the capacity to perceive social signals. Just as a blind person may be aware that he cannot see, but cannot be aware of what he isn't seeing, an Aspie can be aware he doesn't fit in, but lack the ability to perceive just what about him is upsetting others, or how upset they are.
I would advise you to be especially wary of drawing generalized conclusions based upon one annecdotal experience. What may distinguish your friend in your mind is the intensity of his disability, rather than his particular symptoms. But to a trained diagnostician, what distinguishes the Aspie is the particular cluster of symptoms that are characteristic of AS, and of no other kind of social awkwardness.
For example, someone with a social phobia, or PDP NOS will exhibit inability to interact normally with others, but will not display echolalic behaviour, or perserverations, or be unable to recognize faces.
As for what defines someone as an Aspie, we don't know. I don't know, and you don't know either. Until the cause, or causes, and the mechanism, of the phenomenon is understood, all the term can really mean is "We notice that there are a bunch of people who seem to share these traits. It's common enough to remark upon, and to give a name to."
As for just how impaired they are, that's another question. To go back to my example, everyone can see that someone with dark glasses and a white cane is visually impaired. But that doesn't mean that someone with uncorrectible astigmatism is just fine and can learn to see like everyone else. Or that someone with 20/600 vision doesn't need glasses.
Asperger's, like every other disorder, can occur to degrees. And since everyone likes to understand themselves, people can occasionally find it useful to figure out if they display this neurological phenomenon, even if they are not impaired enough to even bother seeking out a diagnosis.
Who said anything about a free pass? Regarding someone as neurologically disordered is hardly a free pass. Hell, the social stigma attached is often far worse than that for being an asshole.
People want to believe that Aspies are fakers, because Aspies generally inspire dislike, which makes people want an excuse for disliking them.
The issue is, if people are really faking, and they *can* be likable, what is it they need an excuse for? Saying that someone is faking Asperger's to have an excuse is a bit like saying someone is faking Tourette's to have an excuse for shouting obscenities in public. If they *didn't* have Tourette's, why would they be shouting them in the first place?
(Because it's a lot more pleasant to fit in than to not fit in, but have an excuse, even if the excuse is accepted.)
You see, I'm not talking about what governments "should" be for. "Should" is really just a way to say "I want" using only one word.
I'm talking about what governments *do*.
Anything which can be done at a profit happens on its own, no government necessary.
Anything else which the society in question wants to do (finds worth doing), it does through a government, whether these actions be enforcing social order, making citizens go to church, exploring space, giving away tax revenues to the company with the most lobbyists, whatever.
In fact, the role of "doing what can't be done at a profit" actually defines government, since a government is that entity in a society which can compel behaviour, and behavioural compulsion is necessary to force people to make investments that do not pay off (i.e., taxes).
When developing a new product or service, there's a most important question to ask.
"What problem does J. Average Person have, that this thing I am selling will solve?"
Doesn't matter whether J. Average Person is supposed to buy the product, or simply use it for free, and allow me to selling advertising. Without bait, no one is going to participate.
So what is it? What's the bait, here? Why do I want to push my data to Google? What problem that I have does this solve?
It sell everything from software to pornography to heating oil, and will immediately either purchase or simply firebomb any competitor out of existence.
And being listed on a known sex offender list IS PART OF THAT PENALTY.
Wrong. It's not being sold as "part of the penalty". If declaring open season on someone was being sold as "part of the penalty", then we're back to death by stoning, right there.
It's being sold as "protection from {$ickybadperson}"
Which of course, is bullshit without clear evidence that it works.
Oh, looky, the plural of "anecdote" has suddenly become "data"!
Look, genius. It's not okay if even one person gets hurt because of this. Not. Okay.
And that's true no matter what he or she has done. Because there is an appropriate penalty for what they have done, and that is to be administered according to law, by a court.
Not by some yahoo like you who thinks that he is qualified to unilaterally judge what other people deserve.
Seriously, solar power is ridiculously overrated. The energy density of sunlight at the earth's surface is simply too low to be practical. Way too much real estate would have to be used to make any realistic amounts of power, and at those scales, upkeep becomes prohibitively expensive.
And how, precisely, do you propose to deliver this power to the earth's surface?
What do you expect from me? I'm autistic myself. (Well, the "formal" diagnosis is Aspergers, but I consider that to be on the Autistic Spectrum.)
I agree. The only difference between Asperger and Kanner autism is language-onset delay. Not very telling, in my view.
It's not unusual for people with Aspergers to have trouble recognizing the "correctness" of behaviour, facial expressions, etc. Sure, it's not universal, either, but it's definitely not a rarity.
Do I care? No. But, then, I'm not built to care about things like that. This isn't an "I can't help it", because that implies it's wrong to be anything other than a highly socially-aware, socially-structured individual.
Idiot Savant is the same thing. It is just a description, no different from "hot", "yellow" or "crispy". Someone might get angry with the words, finding them offensive. That's not my problem. How you choose to understand words is entirely up to you.
You miss my point. I was talking to someone I thought was an NT, in a forum primarily composed of NTs.
NTs think in language. But an effect of this is that the words they customarily use effect the way they think. In other words, if we let NTs continue the use words like "idiot savant", then this will contribute to them continuing to think of autistic people (with savant abilities or otherwise) as basket cases. Riders of the Short Bus. Members of the Helmet Brigade. Twitchy, non-lucid people who must be 'taken care of' rather than understood. You get the idea.
I'm not *personally* offended. Being who and what I am, I attach importance to principles, not individual interactions.
But principle is what I'm talking about here. Twisting a few arms to get people to use "politically correct" language is sometimes appropriate, not because of the meaning (or lack thereof) of words, but because it forces them to acknowledge the existance and humanity of those for whom they must change their behaviour.
Many of the better code-breakers in history were experts or idiot-savants who "specialized" in the structure of information.
You know, I can put up with almost almost all the insults in this thread. That's what I would expect when a bunch of NTs get together and start talking about autism while totally ignorant on the subject. (The ones with autistic relatives may be slightly more clueful.)
But I'm not going to sit by while you toss around the term "idiot-savant".
Who died and made you Grand Lord Definer of Intelligence? To call an autistic person with savant abilities an "idiot-savant" is to call every autistic person without savant abilities just an "idiot".
Just to make it explicit:
1. Not every autistic person has savant abilities.
2. Not every autistic person is nonverbal.
3. Even the ones who *are* nonverbal are not the sort of basket-case you seem to think they are.
4. There are a lot of autistic adults out here who are quite capable of making ourselves understood, and we are sick of being spoken of in this slighting fashion.
That's all. We now return you to your regularly scheduled program of ignorance, flamewars, and cheat-beating contests (otherwise known as "Slashdot.org").
(Oh, and one brief note to those posters with young autistic relatives: I'm sorry it's been tough for you. Believe me, it's no picnic for them, either. Try getting them a computer; keyboards are wonderful things. Don't despair; a significant portion of us *do* learn to communicate your way as we become adults. And don't go for the "dog-training" programs; trust me, they're bad juju...you'd be better off with acceptance, patience, and a degree of flexability.)
You're absolutely right that "spawn camping doesn't scale". However, I think the answer is not instancing. He's right that it's a decent short-term solution. The real solution will be significantly more dynamic and unique content...The genius who develops a world engine bright enough to create and manage AI'd content around one-time quests or events
Don't wait up nights. We are at least decades away from an AI bright enough to have anything resembling creativity, and randomness is *not* an acceptable substitute.
Developing a system bright enough to generate interesting content on its own will not be a function of "genius", but of better hardware.
You see, human "intelligence" isn't actually that sophisticated. In fact, the strategies that humans use to solve many problems are actually *less* sophisticated than the equivalent AI techniques. But the human brain is more powerful by many orders of magnitude than today's best supercomputers.
Now, I agree with you that such a game would have indefinite replay value, and that such a solution would eventually be the desirable one.
But generating interesting stories/game content is a problem that we in the field refer to somewhat jokingly as "AI-complete"; that is to say, any system capable of complex enough behaviour to solve this problem could also be made to pass the Turing test, and would ipso facto be a human being.
Yes. However, I disagree with some of his conclusions.
{snip material on modularization, re-use, high-level languages, "extreme" programming, etc}
Well, most of these have helped, but they are not, and won't be, the quantum leap that would allow us to start narrowing the design productivity gap.
The problem is that hardware capability leaps have resulted from one or two processes, mostly involving shrinking the CMOS transistor. The geometric gains this has produced is the result of literally a geometric factor: A linear reduction in the size of PMOS/NMOS results in a quadratic increase in the amount that will fit on a wafer.
But there is no physical geometric factor in software design. So our linear gains in efficiency remain linear, and our design process capabilities cannot keep up with what our hardware can support.
It's not that software engineering is inherently blighted. It's that hardware design is inherently blessed. And without some unlikely (not impossible, but unlikely) discovery that will somehow bless the software design process in the same manner, the design productivity gap will continue to expand until such time as the compiler itself can read a specification document and write the requested code. (An 'AI-complete' problem.)
Transformation has not really much to do with Deciding whether two programs are equivalent(EQTM).
It has everything to do with it. Two TMs (or programs, or subroutines) which produce the same result for every input are equivalent. Two which don't, aren't.
A compiler may only replace code sequences with equivalent ones. It cannot make this decision itself, because to do so, it would have to solve EQTM.
So just what the hell is it that compilers do?
They match and replace patterns. But the equivalence of these pattern pairs (replace-x-with-y) was predetermined by a human being. And human beings do this heuristically, rather than algorithmically, which is why we can do it at all.
Nevertheless its possible to automate that transformation.
"Can automate" != "algorithmic"
Note: in case your original program would not halt on a given input, your transformed one would neither... this topic is not touched at all.
Please go back and read carefully what I said about one problem reducing to the other. If you are still confused after that, I recommend Michael
Sipser's excellent (and mercifully concise) textbook "Introduction to the Theory of Computation."
Re:you have no idea what you're talking about
on
Java Faster Than C++?
·
· Score: 2, Insightful
Your other assertion here is that a recursive algorithm rewritten iteratively is still recursive if you don't do anything more complicated than maintain the state with a stack. Well this is not true. A function that does not invoke itself is not recursive, period.
Depends how you define recursion. The prescence of a stack which maintains states means that the function *is* invoking itself. It's just doing so in the same frame on the main stack.
Externalizing the stack *is* more effecient. But it is still, in any meaningful sense of the term, recursive.
In fact just by looking at some (non-recursive) function that uses a stack, and judging that it could obviously be rewritten as recursion, you are finding an equivalance between them, which according to you is impossible.
You have misunderstood the meaning of the term "undecideable". Please see my other reply for the details of how this works.
The undecidability if two random programs are equivalent it does not mean that there is no algorithmic way to do the transformation. It would mean that ANY transformation is impossible, which is, of course silly.
No. The key word here is "algorithmic". If we use artificial intelligence (i.e. heuristic search) techniques, we can transform a program to an equivalent one most of the time.
This is essentially what human programmers are doing when they rewrite code.
However, we lose something here; we cannot guarantee equivalence if we do it this way, which is why it is inappropriate for compilers. You wouldn't want a compiler which could make mistakes. Debugging your own code is difficult enough.
It is important to understand just what we mean when we say a problem is "undecideable". It doesn't mean individual instances of the problem cannot be solved. It means that there is no algorithm for solving the problem.
So we cannot insert code into a compiler which can remove recursive functions and replace them with iterative ones. However, we *can* "pre-solve" individual patterns of recursive functions, program the compiler to recognize these patterns, and to replace them with the iterative pattern we code in for it.
I wouldn't charaterize this as a very good idea, however. Most of these replacements involve dynamic programming, which is something any well-trained programmer should be familiar with anyway. It's dangerous to egregiously substitute a tool for a skill.
Re:you have no idea what you're talking about
on
Java Faster Than C++?
·
· Score: 3, Insightful
I'm sorry, but I am afraid you are wrong.
Well, I know the halting problem and I don't see how it relates to what you're saying. The conversion from recursive to iterative isn't arbitrarily complex, it's simple and mechanical. The easiest way is to simply use a stack to maintain the state of what were previously recursive calls.
This is still recursion. You have simply substituted one stack for another.
Anyways, all the halting problem implies is that an optimizer will never be able to find every situation where a particular optimization is applicable.
No, it implies much more than this.
Deciding whether two programs are equivalent(EQTM) is equivalent to ATM (the halting problem). (I'll give the series of proofs if anyone requests it.)
This means that there is no algorithmic way to transform recursive functions into iterative ones, and any substitution must be based upon matching of patterns pre-coded into the compiler.
Perhaps I did not explain myself as well as I should have. I am not advocating turning CS into a craft. What I am advocating is bringing those elements of craftwork that are missing from the CS cirriculum into the mix, while keeping the proper level of math.
Ah, okay, I get you now.
You don't want so much to *change* the CS curiculum as to add more coding experience to it.
I agree *to some extent*. I think that what needs to be taught that isn't is some coursework on structuring software *at a lower-division level*.
The problem is not that CS students are doing too much computer science; it's they are doing too little; specificly, they are doing too little software engineering.
What happens is that in basic CS courses, the ones that teach programming (usually C++), no attention is paid to teaching students *how to structure and plan*, starting with "here's how to psuedocode" and teaching them how to modularize properly, how to design a class interface, etc. I've been advocating this for quite some time, but to little avail, since it would cost money.
Most "software engineering" courses are taught at the upper-division level, where the damage is already done. Practice doesn't make perfect, it makes permanent (which is why I don't think "self-taught" programming is a good idea).
Every discipline has this to a certain extent: English majors learn to write by not only studying grammer, but by writing. Physicists(sic) not only study physics, they perform experiments to tie their observations to the theory behind it. On the same token, Computer Science majors should similarly do more hands on practical work.
Actually, they do far more than other majors in this regard. Far, far more.
But you are correct that some of them are not quite prepared to be programmers. As with any discipline, some are going to graduate with excellent kung fu, while some did just enough actual coding to get by with C's. It's a skills problem, not a techniques problem.
I assume you have read Brooks's "No Silver Bullet"?.
I work for a telecommunications company. I design and build tools for operational users, and do system integration where needed to glue multiple diverse systems together. I am a jack of all trades.
Okay, then I take it you are someone who has taught himself to program, rather than going through a computer science curiculum{sp?}?
This explains some of your views on what mathematical computer science is; you're going off your impressions based on the way a couple of people with BSCS degrees behave.
tight deadlines prevent you from taking careful mathematical analysis of every aspect of a system.
Not really. Time and space complexity analysis should take no more than five minutes for even the slowest of students (and trust me, some of them are dumb as a sack of wet hammers).
What takes time is designing alogrithms with lower time/space complexity (several large subfields of CS are devoted entirely to this activity).
Now, it is a mistake to spend too much time designing/finding the lowest time/space complexity alogrithm when a simpler one will do. But when you say:
they are a strict waterfall lifecycle development shop - and I suspect they do all the mathematical computability and algorithm analysis that you advocate for every module
...it makes me suspect that you do not have to clear an idea what this sort of algorithm analysis actually is. It takes five minutes. Really.
What I think is going on here is that you're frustrated by them spending a lot of time planning and it seems like they're spinning their wheels rather than accomplishing anything.
Now, in your defense, I will say that it's entirely possible that these guys are not doing a very good *job* of planning, and could probably use some time on the whetstone, but *in general*, what you say about planning vs XP reminds me of a joke I once heard:
Foreman: What'cha doin', Louie? Louie: I'm sawin' these planks, boss... got a bunch of 'em to do! Foreman: Ummmmm, yeah... Louie, I can't tell for sure from this angle but it looks like your saw's kinda dull.
Louie: Eyes of a hawk, boss... I seen better edges on butter-knives!
Foreman: Ummmm, yeah... so tell me, Louie, if you don't mind my askin'... why don'cha sharpen the saw, then?
Louie: Oh, I can't do that now, boss... I'm too busy cuttin' these planks!
You see my point. When a system reaches a certain level of complexity, making something that is properly modularized, well-structured, and easily debugged is analougous to sharpening the saw. It takes a lot of time, but the alternative is worse.
I don't have all the answers - but I do recognize a problem when I see it. How would you deal with it?
There's a difference between accidental and essential complexity.
-Accidental complexity is problems that we create on our own and can be fixed. For example, the complexity of writing and optimizing assembly code can be removed by writing programs in C++.
-Essential complexity is caused by the problem to be solved, and nothing can remove it. If users want a program to do 30 different things, then those 30 things are essential and the program must do those 30 different things.
The problem here is that most accidental complexity has already been removed from modern software engineering practices, so there's no magical way to make the rest simpler.
Turning programming back from a science/engineering field into a craft will definitely NOT help.
One of the things a PhD student is supposed to learn is scientific ethics...If graduates go on to be bad scientific citizens, this demonstrates that they have NOT learned what the university was trying to teach them. I believe that's why they feel justified in taking away his PhD.
I still disagree.
If we characterize this as subsequent evidence of a failure to learn ethics (and therefore grounds for PhD removal), then we are implictly accepting that subsequent evidence of having failed to learn some other material (in which they may have had one or no courses, and which may be outside their specific specialization ) is grounds for removal of a PhD.
The only thing that I think justifies the revocation of a PhD is proof of falsified work.
The PhD doesn't certify you as nice, honest, trustworthy, reliable, or intelligent. It doesn't make any statement that you can design, teach, build stuff, or find your ass with both hands and a flashlight. It only certifies that you successfully completed X requirements listed for the piece of paper. That's it.
So unless it turns out that you didn't *really* complete them, there is logically sound reason to think you can 'revoke' one.
Funny, I always learned from the book because the professors didn't really give a damn and certainly didn't have time to explain something to more than a very few people. So yes, I admire the books.. They have always been my greatest teachers..
And the books were written by...?
Oh, yes, that's right, invisible fairies. How silly of me to forget.
Slashdot Mirror
The real Aspie... lacks the capacity to know he doesn't fit in. That's what makes him an Aspie.
Wide of the mark. What can identify someone as an Aspie is that he lacks the capacity to perceive social signals. Just as a blind person may be aware that he cannot see, but cannot be aware of what he isn't seeing, an Aspie can be aware he doesn't fit in, but lack the ability to perceive just what about him is upsetting others, or how upset they are.
I would advise you to be especially wary of drawing generalized conclusions based upon one annecdotal experience. What may distinguish your friend in your mind is the intensity of his disability, rather than his particular symptoms. But to a trained diagnostician, what distinguishes the Aspie is the particular cluster of symptoms that are characteristic of AS, and of no other kind of social awkwardness.
For example, someone with a social phobia, or PDP NOS will exhibit inability to interact normally with others, but will not display echolalic behaviour, or perserverations, or be unable to recognize faces.
As for what defines someone as an Aspie, we don't know. I don't know, and you don't know either. Until the cause, or causes, and the mechanism, of the phenomenon is understood, all the term can really mean is "We notice that there are a bunch of people who seem to share these traits. It's common enough to remark upon, and to give a name to."
As for just how impaired they are, that's another question. To go back to my example, everyone can see that someone with dark glasses and a white cane is visually impaired. But that doesn't mean that someone with uncorrectible astigmatism is just fine and can learn to see like everyone else. Or that someone with 20/600 vision doesn't need glasses.
Asperger's, like every other disorder, can occur to degrees. And since everyone likes to understand themselves, people can occasionally find it useful to figure out if they display this neurological phenomenon, even if they are not impaired enough to even bother seeking out a diagnosis.
Who said anything about a free pass? Regarding someone as neurologically disordered is hardly a free pass. Hell, the social stigma attached is often far worse than that for being an asshole.
This is used against autistic individuals on all levels of normal function.
m l
The general argument goes like this:
"It's okay for us to torture autistics, because anyone who can object isn't a real autistic. Therefore no one objects."
http://www.autistics.org/library/whoisautistic.ht
Unlearn.
Wishful thinking, really.
People want to believe that Aspies are fakers, because Aspies generally inspire dislike, which makes people want an excuse for disliking them.
The issue is, if people are really faking, and they *can* be likable, what is it they need an excuse for? Saying that someone is faking Asperger's to have an excuse is a bit like saying someone is faking Tourette's to have an excuse for shouting obscenities in public. If they *didn't* have Tourette's, why would they be shouting them in the first place?
(Because it's a lot more pleasant to fit in than to not fit in, but have an excuse, even if the excuse is accepted.)
Er... no.
You see, I'm not talking about what governments "should" be for. "Should" is really just a way to say "I want" using only one word.
I'm talking about what governments *do*.
Anything which can be done at a profit happens on its own, no government necessary.
Anything else which the society in question wants to do (finds worth doing), it does through a government, whether these actions be enforcing social order, making citizens go to church, exploring space, giving away tax revenues to the company with the most lobbyists, whatever.
In fact, the role of "doing what can't be done at a profit" actually defines government, since a government is that entity in a society which can compel behaviour, and behavioural compulsion is necessary to force people to make investments that do not pay off (i.e., taxes).
is if we forget about government taking the lead, and focus on capitalism.
You miss the point. Anything which one can make a profit doing, will eventually be done without "us" (whoever that may be) needing to focus on it.
If it's not getting done without government funding, it probably can't be done at a profit (yet).
That's what governments are for; doing that which is worth the expense of doing, but does not directly yield a profit.
>I count 12 required fields where you have to enter data.
>Is this worth throwing a hissy fit over?
And I count one (1) principle at stake.
Which is *always* worth throwing a fit over.
When developing a new product or service, there's a most important question to ask.
"What problem does J. Average Person have, that this thing I am selling will solve?"
Doesn't matter whether J. Average Person is supposed to buy the product, or simply use it for free, and allow me to selling advertising. Without bait, no one is going to participate.
So what is it? What's the bait, here? Why do I want to push my data to Google? What problem that I have does this solve?
...there will be only one company. MergeCo.
It sell everything from software to pornography to heating oil, and will immediately either purchase or simply firebomb any competitor out of existence.
People will call this system a "free market".
And all will be well with the world.
And being listed on a known sex offender list IS PART OF THAT PENALTY.
Wrong. It's not being sold as "part of the penalty". If declaring open season on someone was being sold as "part of the penalty", then we're back to death by stoning, right there.
It's being sold as "protection from {$ickybadperson}"
Which of course, is bullshit without clear evidence that it works.
Oh, looky, the plural of "anecdote" has suddenly become "data"!
Look, genius. It's not okay if even one person gets hurt because of this. Not. Okay.
And that's true no matter what he or she has done. Because there is an appropriate penalty for what they have done, and that is to be administered according to law, by a court.
Not by some yahoo like you who thinks that he is qualified to unilaterally judge what other people deserve.
Seriously, solar power is ridiculously overrated. The energy density of sunlight at the earth's surface is simply too low to be practical. Way too much real estate would have to be used to make any realistic amounts of power, and at those scales, upkeep becomes prohibitively expensive.
And how, precisely, do you propose to deliver this power to the earth's surface?
What do you expect from me? I'm autistic myself. (Well, the "formal" diagnosis is Aspergers, but I consider that to be on the Autistic Spectrum.)
I agree. The only difference between Asperger and Kanner autism is language-onset delay. Not very telling, in my view.
It's not unusual for people with Aspergers to have trouble recognizing the "correctness" of behaviour, facial expressions, etc. Sure, it's not universal, either, but it's definitely not a rarity.
Do I care? No. But, then, I'm not built to care about things like that. This isn't an "I can't help it", because that implies it's wrong to be anything other than a highly socially-aware, socially-structured individual.
Idiot Savant is the same thing. It is just a description, no different from "hot", "yellow" or "crispy". Someone might get angry with the words, finding them offensive. That's not my problem. How you choose to understand words is entirely up to you.
You miss my point. I was talking to someone I thought was an NT, in a forum primarily composed of NTs.
NTs think in language. But an effect of this is that the words they customarily use effect the way they think. In other words, if we let NTs continue the use words like "idiot savant", then this will contribute to them continuing to think of autistic people (with savant abilities or otherwise) as basket cases. Riders of the Short Bus. Members of the Helmet Brigade. Twitchy, non-lucid people who must be 'taken care of' rather than understood. You get the idea.
I'm not *personally* offended. Being who and what I am, I attach importance to principles, not individual interactions.
But principle is what I'm talking about here. Twisting a few arms to get people to use "politically correct" language is sometimes appropriate, not because of the meaning (or lack thereof) of words, but because it forces them to acknowledge the existance and humanity of those for whom they must change their behaviour.
Does that make a little more sense to you?
Many of the better code-breakers in history were experts or idiot-savants who "specialized" in the structure of information.
You know, I can put up with almost almost all the insults in this thread. That's what I would expect when a bunch of NTs get together and start talking about autism while totally ignorant on the subject. (The ones with autistic relatives may be slightly more clueful.)
But I'm not going to sit by while you toss around the term "idiot-savant".
Who died and made you Grand Lord Definer of Intelligence? To call an autistic person with savant abilities an "idiot-savant" is to call every autistic person without savant abilities just an "idiot".
Just to make it explicit:
1. Not every autistic person has savant abilities.
2. Not every autistic person is nonverbal.
3. Even the ones who *are* nonverbal are not the sort of basket-case you seem to think they are.
4. There are a lot of autistic adults out here who are quite capable of making ourselves understood, and we are sick of being spoken of in this slighting fashion.
That's all. We now return you to your regularly scheduled program of ignorance, flamewars, and cheat-beating contests (otherwise known as "Slashdot.org").
(Oh, and one brief note to those posters with young autistic relatives: I'm sorry it's been tough for you. Believe me, it's no picnic for them, either. Try getting them a computer; keyboards are wonderful things. Don't despair; a significant portion of us *do* learn to communicate your way as we become adults. And don't go for the "dog-training" programs; trust me, they're bad juju...you'd be better off with acceptance, patience, and a degree of flexability.)
You're absolutely right that "spawn camping doesn't scale". However, I think the answer is not instancing. He's right that it's a decent short-term solution. The real solution will be significantly more dynamic and unique content...The genius who develops a world engine bright enough to create and manage AI'd content around one-time quests or events
Don't wait up nights. We are at least decades away from an AI bright enough to have anything resembling creativity, and randomness is *not* an acceptable substitute.
Developing a system bright enough to generate interesting content on its own will not be a function of "genius", but of better hardware.
You see, human "intelligence" isn't actually that sophisticated. In fact, the strategies that humans use to solve many problems are actually *less* sophisticated than the equivalent AI techniques. But the human brain is more powerful by many orders of magnitude than today's best supercomputers.
Now, I agree with you that such a game would have indefinite replay value, and that such a solution would eventually be the desirable one.
But generating interesting stories/game content is a problem that we in the field refer to somewhat jokingly as "AI-complete"; that is to say, any system capable of complex enough behaviour to solve this problem could also be made to pass the Turing test, and would ipso facto be a human being.
Yes. However, I disagree with some of his conclusions.
{snip material on modularization, re-use, high-level languages, "extreme" programming, etc}
Well, most of these have helped, but they are not, and won't be, the quantum leap that would allow us to start narrowing the design productivity gap.
The problem is that hardware capability leaps have resulted from one or two processes, mostly involving shrinking the CMOS transistor. The geometric gains this has produced is the result of literally a geometric factor: A linear reduction in the size of PMOS/NMOS results in a quadratic increase in the amount that will fit on a wafer.
But there is no physical geometric factor in software design. So our linear gains in efficiency remain linear, and our design process capabilities cannot keep up with what our hardware can support.
It's not that software engineering is inherently blighted. It's that hardware design is inherently blessed. And without some unlikely (not impossible, but unlikely) discovery that will somehow bless the software design process in the same manner, the design productivity gap will continue to expand until such time as the compiler itself can read a specification document and write the requested code. (An 'AI-complete' problem.)
Transformation has not really much to do with Deciding whether two programs are equivalent(EQTM).
... this topic is not touched at all.
It has everything to do with it. Two TMs (or programs, or subroutines) which produce the same result for every input are equivalent. Two which don't, aren't.
A compiler may only replace code sequences with equivalent ones. It cannot make this decision itself, because to do so, it would have to solve EQTM.
So just what the hell is it that compilers do?
They match and replace patterns. But the equivalence of these pattern pairs (replace-x-with-y) was predetermined by a human being. And human beings do this heuristically, rather than algorithmically, which is why we can do it at all.
Nevertheless its possible to automate that transformation.
"Can automate" != "algorithmic"
Note: in case your original program would not halt on a given input, your transformed one would neither
Please go back and read carefully what I said about one problem reducing to the other. If you are still confused after that, I recommend Michael Sipser's excellent (and mercifully concise) textbook "Introduction to the Theory of Computation."
Your other assertion here is that a recursive algorithm rewritten iteratively is still recursive if you don't do anything more complicated than maintain the state with a stack. Well this is not true. A function that does not invoke itself is not recursive, period.
Depends how you define recursion. The prescence of a stack which maintains states means that the function *is* invoking itself. It's just doing so in the same frame on the main stack.
Externalizing the stack *is* more effecient. But it is still, in any meaningful sense of the term, recursive.
In fact just by looking at some (non-recursive) function that uses a stack, and judging that it could obviously be rewritten as recursion, you are finding an equivalance between them, which according to you is impossible.
You have misunderstood the meaning of the term "undecideable". Please see my other reply for the details of how this works.
The undecidability if two random programs are equivalent it does not mean that there is no algorithmic way to do the transformation. It would mean that ANY transformation is impossible, which is, of course silly.
No. The key word here is "algorithmic". If we use artificial intelligence (i.e. heuristic search) techniques, we can transform a program to an equivalent one most of the time.
This is essentially what human programmers are doing when they rewrite code.
However, we lose something here; we cannot guarantee equivalence if we do it this way, which is why it is inappropriate for compilers. You wouldn't want a compiler which could make mistakes. Debugging your own code is difficult enough.
It is important to understand just what we mean when we say a problem is "undecideable". It doesn't mean individual instances of the problem cannot be solved. It means that there is no algorithm for solving the problem.
So we cannot insert code into a compiler which can remove recursive functions and replace them with iterative ones. However, we *can* "pre-solve" individual patterns of recursive functions, program the compiler to recognize these patterns, and to replace them with the iterative pattern we code in for it.
I wouldn't charaterize this as a very good idea, however. Most of these replacements involve dynamic programming, which is something any well-trained programmer should be familiar with anyway. It's dangerous to egregiously substitute a tool for a skill.
I'm sorry, but I am afraid you are wrong.
Well, I know the halting problem and I don't see how it relates to what you're saying. The conversion from recursive to iterative isn't arbitrarily complex, it's simple and mechanical. The easiest way is to simply use a stack to maintain the state of what were previously recursive calls.
This is still recursion. You have simply substituted one stack for another.
Anyways, all the halting problem implies is that an optimizer will never be able to find every situation where a particular optimization is applicable.
No, it implies much more than this.
Deciding whether two programs are equivalent(EQTM) is equivalent to ATM (the halting problem). (I'll give the series of proofs if anyone requests it.)
This means that there is no algorithmic way to transform recursive functions into iterative ones, and any substitution must be based upon matching of patterns pre-coded into the compiler.
Spammers are people, and people can have morals, therefore spammers can have morals. Why not?
Armadillos are animals, and animals can have wings. Therefore, armadillos can have wings.
Perhaps I did not explain myself as well as I should have. I am not advocating turning CS into a craft. What I am advocating is bringing those elements of craftwork that are missing from the CS cirriculum into the mix, while keeping the proper level of math.
Ah, okay, I get you now.
You don't want so much to *change* the CS curiculum as to add more coding experience to it.
I agree *to some extent*. I think that what needs to be taught that isn't is some coursework on structuring software *at a lower-division level*.
The problem is not that CS students are doing too much computer science; it's they are doing too little; specificly, they are doing too little software engineering.
What happens is that in basic CS courses, the ones that teach programming (usually C++), no attention is paid to teaching students *how to structure and plan*, starting with "here's how to psuedocode" and teaching them how to modularize properly, how to design a class interface, etc. I've been advocating this for quite some time, but to little avail, since it would cost money.
Most "software engineering" courses are taught at the upper-division level, where the damage is already done. Practice doesn't make perfect, it makes permanent (which is why I don't think "self-taught" programming is a good idea).
Every discipline has this to a certain extent: English majors learn to write by not only studying grammer, but by writing. Physicists(sic) not only study physics, they perform experiments to tie their observations to the theory behind it. On the same token, Computer Science majors should similarly do more hands on practical work.
Actually, they do far more than other majors in this regard. Far, far more.
But you are correct that some of them are not quite prepared to be programmers. As with any discipline, some are going to graduate with excellent kung fu, while some did just enough actual coding to get by with C's. It's a skills problem, not a techniques problem.
I assume you have read Brooks's "No Silver Bullet"?.
I work for a telecommunications company. I design and build tools for operational users, and do system integration where needed to glue multiple diverse systems together. I am a jack of all trades.
...it makes me suspect that you do not have to clear an idea what this sort of algorithm analysis actually is. It takes five minutes. Really.
Okay, then I take it you are someone who has taught himself to program, rather than going through a computer science curiculum{sp?}?
This explains some of your views on what mathematical computer science is; you're going off your impressions based on the way a couple of people with BSCS degrees behave.
tight deadlines prevent you from taking careful mathematical analysis of every aspect of a system.
Not really. Time and space complexity analysis should take no more than five minutes for even the slowest of students (and trust me, some of them are dumb as a sack of wet hammers).
What takes time is designing alogrithms with lower time/space complexity (several large subfields of CS are devoted entirely to this activity).
Now, it is a mistake to spend too much time designing/finding the lowest time/space complexity alogrithm when a simpler one will do. But when you say:
they are a strict waterfall lifecycle development shop - and I suspect they do all the mathematical computability and algorithm analysis that you advocate for every module
What I think is going on here is that you're frustrated by them spending a lot of time planning and it seems like they're spinning their wheels rather than accomplishing anything.
Now, in your defense, I will say that it's entirely possible that these guys are not doing a very good *job* of planning, and could probably use some time on the whetstone, but *in general*, what you say about planning vs XP reminds me of a joke I once heard:
Foreman: What'cha doin', Louie?
Louie: I'm sawin' these planks, boss... got a bunch of 'em to do!
Foreman: Ummmmm, yeah... Louie, I can't tell for sure from this angle but it looks like your saw's kinda dull.
Louie: Eyes of a hawk, boss... I seen better edges on butter-knives!
Foreman: Ummmm, yeah... so tell me, Louie, if you don't mind my askin'... why don'cha sharpen the saw, then?
Louie: Oh, I can't do that now, boss... I'm too busy cuttin' these planks!
You see my point. When a system reaches a certain level of complexity, making something that is properly modularized, well-structured, and easily debugged is analougous to sharpening the saw. It takes a lot of time, but the alternative is worse.
I don't have all the answers - but I do recognize a problem when I see it. How would you deal with it?
There's a difference between accidental and essential complexity.
-Accidental complexity is problems that we create on our own and can be fixed. For example, the complexity of writing and optimizing assembly code can be removed by writing programs in C++.
-Essential complexity is caused by the problem to be solved, and nothing can remove it. If users want a program to do 30 different things, then those 30 things are essential and the program must do those 30 different things.
The problem here is that most accidental complexity has already been removed from modern software engineering practices, so there's no magical way to make the rest simpler.
Turning programming back from a science/engineering field into a craft will definitely NOT help.
One of the things a PhD student is supposed to learn is scientific ethics...If graduates go on to be bad scientific citizens, this demonstrates that they have NOT learned what the university was trying to teach them. I believe that's why they feel justified in taking away his PhD.
I still disagree.
If we characterize this as subsequent evidence of a failure to learn ethics (and therefore grounds for PhD removal), then we are implictly accepting that subsequent evidence of having failed to learn some other material (in which they may have had one or no courses, and which may be outside their specific specialization ) is grounds for removal of a PhD.
The only thing that I think justifies the revocation of a PhD is proof of falsified work.
The PhD doesn't certify you as nice, honest, trustworthy, reliable, or intelligent. It doesn't make any statement that you can design, teach, build stuff, or find your ass with both hands and a flashlight. It only certifies that you successfully completed X requirements listed for the piece of paper. That's it.
So unless it turns out that you didn't *really* complete them, there is logically sound reason to think you can 'revoke' one.
Funny, I always learned from the book because the professors didn't really give a damn and certainly didn't have time to explain something to more than a very few people. So yes, I admire the books.. They have always been my greatest teachers..
And the books were written by...?
Oh, yes, that's right, invisible fairies. How silly of me to forget.