Enjoy the math. Math is your friend! You know by now, I'm sure, that it will help you, not only with direct skills but with additional training in critical thinking.
If you're not enjoying your math classes, please allow me to suggest a few that I've really enjoyed, and that you might enjoy and find helpful as a CS person (in decending order): * General discrete math * Combinatorics * Number theory (a special case of...) * Abstract algebra
All of these things are applicable to computer science (none of them involve real numbers;) ), and are incredibly helpful in terms of logic and foundational math for a large part of CS.
>If you wanted the comic book, you should have bought it when it came out. If you weren't interested when it came out, tough shit. Someone else was and they beat you to it.
I'm not taking a position on this, but I'll just remark that what you said can sound like this:
"If you wanted some land/money/food, you should have bought it when it was available. If you weren't alive when it came out, tough shit. Someone else beat you to it, feel free to be poor."
All I'm saying is that feudelism is generally not considered a useful form of government... that argument is not a generally good one in either case.
The grandparent poster was asking about the dissemination of information, not about the widgets themselves. Reprints of the Comedia, for example, are much, MUCH less expensive than anything even close to an original copy by a few centuries.
I hear that studded tires don't work all that well when you've got a mix of ice, snow, and pavement, but I haven't tried em. Pittsburgh is frankly scary enough getting around in the winter without the slip factor on my bike.
The other really big killer factor in the winter are HUGE GAPING potholes. One ate my car wheel, and would have eaten my bike (and me) alive. (Swerving was not an option with this one -- it had the entire lane and there was oncoming traffic.)
testing this sort of thing is very nontrivial. when you're talking about cutting-edge research, often there are only a small number of people who understand that research area and its implications. you generally want someone who works in that area to do the review... testing for knowledge in every little subarea is not exactly a productive use of someone's time.
Err... actually, that's more of a language theorist's research than a security researcher's work. There are other interesting things that security people do, like provide secure ways to do things which weren't possible before, reduce the amount of trust we have to have in our own or other people's computers (and other people, in certain circumstances), and allow private transactions of all sorts to take place.
The problem with encouraging "any one willing" to review papers is that they might well not understand whether the paper under consideration is a) original b) correct c) interesting (the test for this for conferences in my area has been described as "if the committe can't come up with a solution at least as good in about 15 minutes, it's sufficiently hard work")
Even with the current system, there are some rather egregious failures in both directions, as well as many smaller ones. I once had a paper review that more or less called I and my coauthors crackpots, when in fact it was a simple misunderstanding of the standard meaning of the term we were using (which we defined in italics to try to avoid this problem).
In cryptography there are in fact a fair number of fairly... interesting papers submitted by people with a lack of understanding of the mathematics involved and an axe to grind. Having these people with insufficient background reviewing papers is not likely to help.
Despite it having the highest failure rate of any AP exam (at least 7 years ago when I took it), the CA AB AP exam is pretty much a joke if you have a basic understanding of data structures. However, this is the idea -- if you do well on this, then you don't need to take data structures as an undergrad (and I didn't).
Nearly my entire class finished the multiple choice part in ~30 minutes, and the programming part in ~20, IIRC, and since they made us spend the rest of the time there, we spent the rest of it waving at each other and pissing off the proctors.It's an easy test, probably with a pretty strict division into two bell-curves: has a clue and has no clue.
I took calc at my local high school (which is a very good school, I have to admit) and skipped two semesters of calc at Berkeley. (I had quite a few friends at MIT, so I know what's in that course as well) My point is simply that all calculus courses are not created equal, just as you said, but "lower level" courses aren't necessarily less intense or less useful. I found my high school education served me quite well.
>To explain mysef, relatively new programming languages are all just context-free grammars...
In addition to this, you might also want to note that programming langauges are also heavily invested in type theory, which is another interesting area of computer science.
Possibily the most important part of automota theory is decidability. It is/extremely/ important to understand what computers can and cannot do. It also leads to not a little amusement on the part of theorists at the expense of more "practical" people, who fail to understand that what they are attempting is provably impossible.
There are, though, absolute proofs that use probability. There is a technique that I believe was pioneered by Erdos, where one takes a random graph on a certain number of nodes, and proves that the probability of a certain characteristic (or the probability that if it doesn't have it, it can be manipulated to do so) is either 0 or 1. Thus, all graphs of that size have the characteristic (or don't have it), but proving it the other way around is a real pain. Random graphs are a useful tool, and they do give you an unequivical answer in this case.
You make some very good points, I just wanted to comment on one thing:
>Look into history: Spain was once occupied by the (muslim) Moors; under their reign christians as well as jews could openly have their own religion.
Under the Moors, the Jews were allowed practice of their religion and so forth. Under the Christians, we had the Spanish Inquision, and mass slaughter of Jews. If we're going for interolerance, in this case the Christians come out quite far ahead. Of course, there are examples to the contrary as well.
The US has started getting a lot better about this, apparently, in the face of quite a few countries that do not allow revokation of citizenship. The new standard is "did you intend to revoke your US citizenship" by doing whatever (for example, getting a new citizenship, including loyalty oath). The standard is that they just ask you, though if you DO wish to revoke it, you should go to the US consulate and tell em that and get it recorded. It's not a bad idea to tell them if you don't, either (though a notarized letter will suffice).
And yes, the reccomended practice is that you use whatever passport is appropriate. If you're in Canada, and have a Canadian citizenship, they have dominion over you as a citizen, what do they care that you're also Syrian or Chinese or American or whatever (for example)?
I'm sorry, I'm not really "obviously" anything, other than A THEORETICAL CRYPTOGRAPHER. I did my undergrad at UC Berkeley, not at CMU. Note that my former field of research was robotics, so I actually have gotten my hands extremely dirty. I simply prefer math, and coding annoys me.
Any serious algorithm development effort needs:
A better result than the one known before. This can mean deveoping new capabilities that were not known before, an asymptotically better running time (and note that the cost function may or may not take into account such things as a cache hierarchy, but tying it to a specific platform makes it rather useless, in the long run), a weaker model (fewer assumptions), or something of the sort.
If you are using DATA to prove your algorithms, you're not generally proving them (yes, exhaustive search may be able to prove the desired properties, but see the lack of useful theorem provers for a motivation for actual proofs). A proof guarentees correctness (and whatever other properties you wish, such as security) in a given model. It takes skill, of course, to understand and implement the proved algorithm/protocol. Security is not an easy endevour, for instance; many examples of faulty implementation (such as WEP) show this. The most useful scientists are those who can understand the math. A background in programming is also helpful, but languages can be picked up as needed.
Frankly, if you're using C to prove your algorithms, I'm either very scared, or you have some people using some incredible automatic theorem-provers. Still, I hope you're not rolling your own crypto. Attitudes like that have led to some incredibly silly security flaws.
I actually did my undergrad at UC Berkeley, I'm a PhD candidate at CMU. I didn't have any trouble passing my classes that required programming, actually, but I find it very annoying and choose not to do it often. Once you fix a proof, it STAYS fixed.
programming is not the be-all and end-all of computer science. I know some brilliant theorists who can't code well, but have made startling contributions to the field (and to your daily life).
but hey, all the crypto people wear black leather and sunglasses all the time, so it's ok if they don't program, since they're cool.
You just said it's dependent, not independent. If it's independent, the conditional distribution of belief-in-god given great-thinker the same as the distribution of belief-in-god.
No one's succesfully argued either one in this thread, and it's rather a silly argument in any case.
> Regardless of your opinion on rights and what not, you have to agree that a law that says a union between homosexuals should not be called "marriage" will alienate the homosexuals no less than saying that "marriage" can refer to homosexual unions will alienate those people who believe that is just horrendous.
I believe that for many people, it's very much about the rights associated with marriage. In any case, it's something that changes for homosexuals, not heterosexuals, so in some sense it is more discriminatory to deny the label than to allow it.
And the "keep your way of tought out of this" that you're referring to is very much a "this is not the appropritate forum for an argument of that sort".
Based on their experience of war from not-so-long below, the French acted quite intelligently. The Maginot Line was not quite up to the whole tank thing, but wasn't a bad idea in the grander scheme of things.
There are certainly quite a few wars that the US has lost as well. Several incursions against Mexico (rather slaughter-happy ones) turned out badly, an attack on the USSR which managed to piss them off quite thoroughly, several attacks on Canada (which was at the time a British property), North Korea, and so on and so forth. This is (of course) not counting the minor covert operations, though many of them did not fare well either.
Of course, military history is not my area of specialty, but I thought I might as well point out that there are certainly two sides to this argument.
On that note, I'm frankly amazed by the anti-French military jokes. The French have won some amazing victories, and been enormously powerful militarily. Remember Napolean (who was technically Corsican, but no matter)?
Slashdot Mirror
Enjoy the math. Math is your friend! You know by now, I'm sure, that it will help you, not only with direct skills but with additional training in critical thinking.
;) ), and are incredibly helpful in terms of logic and foundational math for a large part of CS.
:)
If you're not enjoying your math classes, please allow me to suggest a few that I've really enjoyed, and that you might enjoy and find helpful as a CS person (in decending order):
* General discrete math
* Combinatorics
* Number theory (a special case of...)
* Abstract algebra
All of these things are applicable to computer science (none of them involve real numbers
Hope you enjoy em
Lea
>If you wanted the comic book, you should have bought it when it came out. If you weren't interested when it came out, tough shit. Someone else was and they beat you to it.
I'm not taking a position on this, but I'll just remark that what you said can sound like this:
"If you wanted some land/money/food, you should have bought it when it was available. If you weren't alive when it came out, tough shit. Someone else beat you to it, feel free to be poor."
All I'm saying is that feudelism is generally not considered a useful form of government... that argument is not a generally good one in either case.
The grandparent poster was asking about the dissemination of information, not about the widgets themselves. Reprints of the Comedia, for example, are much, MUCH less expensive than anything even close to an original copy by a few centuries.
Lea
I hear that studded tires don't work all that well when you've got a mix of ice, snow, and pavement, but I haven't tried em. Pittsburgh is frankly scary enough getting around in the winter without the slip factor on my bike.
The other really big killer factor in the winter are HUGE GAPING potholes. One ate my car wheel, and would have eaten my bike (and me) alive. (Swerving was not an option with this one -- it had the entire lane and there was oncoming traffic.)
This is generally true of most disciplines, even the humanities, from what I've been told.
Lea
testing this sort of thing is very nontrivial. when you're talking about cutting-edge research, often there are only a small number of people who understand that research area and its implications. you generally want someone who works in that area to do the review... testing for knowledge in every little subarea is not exactly a productive use of someone's time.
Lea
Err... actually, that's more of a language theorist's research than a security researcher's work. There are other interesting things that security people do, like provide secure ways to do things which weren't possible before, reduce the amount of trust we have to have in our own or other people's computers (and other people, in certain circumstances), and allow private transactions of all sorts to take place.
Lea
The problem with encouraging "any one willing" to review papers is that they might well not understand whether the paper under consideration is a) original b) correct c) interesting (the test for this for conferences in my area has been described as "if the committe can't come up with a solution at least as good in about 15 minutes, it's sufficiently hard work")
... interesting papers submitted by people with a lack of understanding of the mathematics involved and an axe to grind. Having these people with insufficient background reviewing papers is not likely to help.
Even with the current system, there are some rather egregious failures in both directions, as well as many smaller ones. I once had a paper review that more or less called I and my coauthors crackpots, when in fact it was a simple misunderstanding of the standard meaning of the term we were using (which we defined in italics to try to avoid this problem).
In cryptography there are in fact a fair number of fairly
Lea
Hmm... I think one of you must be lying, or the count is somewhat off, as I certainly reside in this category as well.
Of course, like the good sterotypical geek woman some of you will imagine to be, I've been spoken for.
Lea
Many schools do not allow TAs to teach classes. However, this is still rather unreasonable for the students, especially the ones who have graduated.
Lea
Are you telling me that moving the decimal place back and forth is HARDER than dividing by, say, 3?
*boggle*
Lea
Try -O2. I believe O3 includes some optimizations which generally make things run faster, but not always. Aggressive optimization is a chancy thing.
Lea
Despite it having the highest failure rate of any AP exam (at least 7 years ago when I took it), the CA AB AP exam is pretty much a joke if you have a basic understanding of data structures. However, this is the idea -- if you do well on this, then you don't need to take data structures as an undergrad (and I didn't).
Nearly my entire class finished the multiple choice part in ~30 minutes, and the programming part in ~20, IIRC, and since they made us spend the rest of the time there, we spent the rest of it waving at each other and pissing off the proctors.It's an easy test, probably with a pretty strict division into two bell-curves: has a clue and has no clue.
Lea
I don't know if you've seen this, but there's actualyl quite a lot of work on variants of undecidable problems which are themselves decidable.
Redefining the problem is not the same as solving it. Both are important.
Lea
I took calc at my local high school (which is a very good school, I have to admit) and skipped two semesters of calc at Berkeley. (I had quite a few friends at MIT, so I know what's in that course as well) My point is simply that all calculus courses are not created equal, just as you said, but "lower level" courses aren't necessarily less intense or less useful. I found my high school education served me quite well.
Lea
>To explain mysef, relatively new programming languages are all just context-free grammars...
/extremely/ important to understand what computers can and cannot do. It also leads to not a little amusement on the part of theorists at the expense of more "practical" people, who fail to understand that what they are attempting is provably impossible.
In addition to this, you might also want to note that programming langauges are also heavily invested in type theory, which is another interesting area of computer science.
Possibily the most important part of automota theory is decidability. It is
Lea
There are, though, absolute proofs that use probability. There is a technique that I believe was pioneered by Erdos, where one takes a random graph on a certain number of nodes, and proves that the probability of a certain characteristic (or the probability that if it doesn't have it, it can be manipulated to do so) is either 0 or 1. Thus, all graphs of that size have the characteristic (or don't have it), but proving it the other way around is a real pain. Random graphs are a useful tool, and they do give you an unequivical answer in this case.
Lea
You make some very good points, I just wanted to comment on one thing:
>Look into history: Spain was once occupied by the (muslim) Moors; under their reign christians as well as jews could openly have their own religion.
Under the Moors, the Jews were allowed practice of their religion and so forth. Under the Christians, we had the Spanish Inquision, and mass slaughter of Jews. If we're going for interolerance, in this case the Christians come out quite far ahead. Of course, there are examples to the contrary as well.
Lea
The US has started getting a lot better about this, apparently, in the face of quite a few countries that do not allow revokation of citizenship. The new standard is "did you intend to revoke your US citizenship" by doing whatever (for example, getting a new citizenship, including loyalty oath). The standard is that they just ask you, though if you DO wish to revoke it, you should go to the US consulate and tell em that and get it recorded. It's not a bad idea to tell them if you don't, either (though a notarized letter will suffice).
And yes, the reccomended practice is that you use whatever passport is appropriate. If you're in Canada, and have a Canadian citizenship, they have dominion over you as a citizen, what do they care that you're also Syrian or Chinese or American or whatever (for example)?
Lea
Any serious algorithm development effort needs:
Frankly, if you're using C to prove your algorithms, I'm either very scared, or you have some people using some incredible automatic theorem-provers. Still, I hope you're not rolling your own crypto. Attitudes like that have led to some incredibly silly security flaws.
Lea
I actually did my undergrad at UC Berkeley, I'm a PhD candidate at CMU. I didn't have any trouble passing my classes that required programming, actually, but I find it very annoying and choose not to do it often. Once you fix a proof, it STAYS fixed.
Lea
programming is not the be-all and end-all of computer science. I know some brilliant theorists who can't code well, but have made startling contributions to the field (and to your daily life).
but hey, all the crypto people wear black leather and sunglasses all the time, so it's ok if they don't program, since they're cool.
Lea
You just said it's dependent, not independent. If it's independent, the conditional distribution of belief-in-god given great-thinker the same as the distribution of belief-in-god.
No one's succesfully argued either one in this thread, and it's rather a silly argument in any case.
Lea
> Regardless of your opinion on rights and what not, you have to agree that a law that says a union between homosexuals should not be called "marriage" will alienate the homosexuals no less than saying that "marriage" can refer to homosexual unions will alienate those people who believe that is just horrendous.
I believe that for many people, it's very much about the rights associated with marriage. In any case, it's something that changes for homosexuals, not heterosexuals, so in some sense it is more discriminatory to deny the label than to allow it.
And the "keep your way of tought out of this" that you're referring to is very much a "this is not the appropritate forum for an argument of that sort".
Lea
Based on their experience of war from not-so-long below, the French acted quite intelligently. The Maginot Line was not quite up to the whole tank thing, but wasn't a bad idea in the grander scheme of things.
Lea
There are certainly quite a few wars that the US has lost as well. Several incursions against Mexico (rather slaughter-happy ones) turned out badly, an attack on the USSR which managed to piss them off quite thoroughly, several attacks on Canada (which was at the time a British property), North Korea, and so on and so forth. This is (of course) not counting the minor covert operations, though many of them did not fare well either.
Of course, military history is not my area of specialty, but I thought I might as well point out that there are certainly two sides to this argument.
On that note, I'm frankly amazed by the anti-French military jokes. The French have won some amazing victories, and been enormously powerful militarily. Remember Napolean (who was technically Corsican, but no matter)?
Lea