So what's your point? Do you want to make the ArXiv a peer-review journal (which kills its uniqueness)? Do you think it should be eliminated? Or are you just complaining for the hell of it?
That's fine if you have the expertise to critically review the papers yourself but if you don't you have to keep in mind that nobody else has done so for you.
Why would anyone read a research paper uncritically? I think it's understood that when you reach the level of expertise required to read research articles, you always read them critically. You criticize the ArXiv as if it were a replacement for peer-reviewed journals. I've never met anyone who thinks this, and it was never meant to be. That is not a bad thing. The problems you mention are not problems at all.
Peer-reviewed papers do commonly have errors. The cool thing about the ArXiv is you can make revisions anytime you want. The best thing about the ArXiv, from a researchers perspective, is that you don't have to wait 2-3 years to see publications on the state of the art.
Even though anybody can put a paper on the ArXiv, there is an incentive to get things right. Take a look at this retraction of a proof of the Riemann hypothesis.
Notice that a link to the original article is given at the bottom. You can't remove submissions from the ArXiv, so whenever you submit an article, you risk having your errors online forever, which may be a good or bad thing, but it certainly makes people more careful.
Not exactly. At least in math, almost everybody posts to the ArXiv now. Some journals allow you to submit papers for publication directly from the ArXiv. Grigori Perelman posted his papers (Poincare conjecture) on the ArXiv years ago. It's true that anything that reasonably looks like a research paper will get accepted into the ArXiv, but anything incorrect that draws attention is usually refuted. Moreover, you can't remove submissions from the ArXiv (but you can revise/retract), which provides an incentive to do careful research. There are a few other proofs of the Riemann hypothesis on the ArXiv (I recall one of only a few pages that was a fun read years ago). I wonder why people are so drawn to it, since it seems to be the hardest of the millenium problems.
You're kidding, right? There is no such thing as an unemployed applied math Ph.D., unless it's by choice. Applied math PhDs can get jobs in almost any field outside academia, like actuary, gov't think tank, private think tank, finance, defense contractor, programmer, etc... I know a lot of math PhDs, and applied math PhDs aren't the ones struggling to find jobs.
Just about every school does this, dude, maybe except Harvard. It's good to know you intend to spend $50,000/year on your kids' education, since no public school in this country (yes, that includes the U of CA schools) produces only competent programmers.
He/she probably decided to get a CS degree before entering college. Once you complete enough units toward your major, it might be easier to work hard and finish on time instead of switching majors and taking another year or more to finish. Or, maybe it was one of the hardest CS programs in the country, and their worst programmers are just average in the industry.
Actually, out in certain suburbs, housing prices have taken a nosedive. Check craigslist for prices in the Inland Empire (60 miles east of Los Angeles) and you'll find several houses for under $200,000 (Moreno Valley, Temecula, etc...)
This isn't true anymore. A 2BD/1BA house in Los Angeles rents for not less than $2000/month. A $400,000 mortgage at 5% is about the same. Right now you can only get a house in the cheapest nbhds for that price, but other nbhds are about to follow.
I'm the same here in Los Angeles. I had a chest CT done, and the doctor wondered if I were a smoker (I'm not). Several years of inhaling dust isn't good for you. I should start vacationing in Seattle:)
It's obvious to me now that you a) have a chip on your shoulder with regards to Muslims (irony in its purest form) and b) know very few or perhaps even none.
a) Of course I do, but how is that ironic?
b) There are over one billion Muslims in the world. What does it matter if you personally know a couple of them? Also, what would you say if I stated that most Muslims are violent thugs because I know a few Muslims who are violent thugs?
But you don't know any, do you? And that's the point. Ignorance.
You'll do even better on file size if you convert scans to Djvu: Scan to pnm (or pgm or pbm), convert using c44 (or cjb2 for pbm), and you can skip the TeX step (use djvm to assemble pages). If the text you're scanning is black and white, you'll be stunned by the quality and file size reduction.
I don't think 300dpi is high for archival purposes. I think it's around the minimum. I can convert handwritten bitonal 8x11in 450dpi scans into 30-40 kilobyte Djvu files.
I bought a used car (private party) with 666 in the license plate. After a relative begged me to swap it, I went to the DMV (California) to change it, since it was only $8. After filling out some papers, the clerk gave me a new plate, and as I was about to leave, I realized I hadn't paid. So I told him, and he said that anyone with a 666 in their license can exchange it for free. I was pleased.
That's interesting. Here in L.A., when there are strong winds, the smog disappears and you can see the mountains. It's really beautiful. A few days later the haze is as thick as it was before. I always wondered where the fog disappears to. So I guess it rises out of view when it's windy, and falls back into place later? That seems strange.
Nice pictures! I also found this quote interesting:
A lovely gin and tonic to start the flight off and the option to stretch ones legs out and work on the MacBook Air in-between looking out the window to take pictures makes every flight much more pleasant.
Is this how people usually turn out when they buy a Mac?:)
Not sure if Windows 2008 is rock solid, but I like it because... it's a free download, so I run it as a KVM virtual machine. Add to that Visual C++ Express Edition, and I don't have to purchase any MS software when I need it for teaching. Pretty cool.
If 90% of your code is math, how about Lisp or Fortran? Maybe Lisp is too slow and it's hard to represent objects in Fortran. Imagine how you would represent a multivariate polynomial with rational coefficients in Fortran.
Altogether, I'd rather do the kind of work you're talking about in Matlab or Octave. Or you could do something like Beanshell -- have an interpreted language that is closely tied to the underlying language and its libraries. Matlab (and its clones) are mostly useless in the numerous areas of math outside of a standard undergrad curriculum, even for "profiling" since the data structures aren't even in the language. It's a zillion times easier to write these in C++. This never gets mentioned on forums like this, but C++ provides the right balance of power and flexibility (especially with memory managament) for people who do specialized math/science. And even in not-so-specialized areas, sometimes objects in Matlab clones are represented in a way which is cumbersome to use. Writing a representation in C++ might save hours.
The good thing about a tax on obesity is that the gov't would be sued into providing a proper definition. Once it's showed that it is common for healthy men of the same height and same body fat percentage to differ in weight by as much as 60 pounds, the commonly-used nonsensical definition of obesity might go away.
Even with private health insurance, those who live unhealthy lifestyles have the net effect of increasing insurance premiums for everyone.
I think history, and expecially recent history, shows this reasoning is not true. The demand for health insurance is inelastic, just as it is for oil. Companies don't have to collude to keep prices high when they know certain helth services are not optional-- they just charge as much as possible, and will charge more in the future.
Slashdot Mirror
So what's your point? Do you want to make the ArXiv a peer-review journal (which kills its uniqueness)? Do you think it should be eliminated? Or are you just complaining for the hell of it?
That's fine if you have the expertise to critically review the papers yourself but if you don't you have to keep in mind that nobody else has done so for you.
Why would anyone read a research paper uncritically? I think it's understood that when you reach the level of expertise required to read research articles, you always read them critically. You criticize the ArXiv as if it were a replacement for peer-reviewed journals. I've never met anyone who thinks this, and it was never meant to be. That is not a bad thing. The problems you mention are not problems at all.
Peer-reviewed papers do commonly have errors. The cool thing about the ArXiv is you can make revisions anytime you want. The best thing about the ArXiv, from a researchers perspective, is that you don't have to wait 2-3 years to see publications on the state of the art.
Even though anybody can put a paper on the ArXiv, there is an incentive to get things right. Take a look at this retraction of a proof of the Riemann hypothesis.
http://arxiv.org/abs/math/0109072
Notice that a link to the original article is given at the bottom. You can't remove submissions from the ArXiv, so whenever you submit an article, you risk having your errors online forever, which may be a good or bad thing, but it certainly makes people more careful.
Not exactly. At least in math, almost everybody posts to the ArXiv now. Some journals allow you to submit papers for publication directly from the ArXiv. Grigori Perelman posted his papers (Poincare conjecture) on the ArXiv years ago. It's true that anything that reasonably looks like a research paper will get accepted into the ArXiv, but anything incorrect that draws attention is usually refuted. Moreover, you can't remove submissions from the ArXiv (but you can revise/retract), which provides an incentive to do careful research. There are a few other proofs of the Riemann hypothesis on the ArXiv (I recall one of only a few pages that was a fun read years ago). I wonder why people are so drawn to it, since it seems to be the hardest of the millenium problems.
You're kidding, right? There is no such thing as an unemployed applied math Ph.D., unless it's by choice. Applied math PhDs can get jobs in almost any field outside academia, like actuary, gov't think tank, private think tank, finance, defense contractor, programmer, etc... I know a lot of math PhDs, and applied math PhDs aren't the ones struggling to find jobs.
You pay $500/month in property taxes outside of California? Where are you? We are talking about a 30 year mortgage, right?
Just about every school does this, dude, maybe except Harvard. It's good to know you intend to spend $50,000/year on your kids' education, since no public school in this country (yes, that includes the U of CA schools) produces only competent programmers.
He/she probably decided to get a CS degree before entering college. Once you complete enough units toward your major, it might be easier to work hard and finish on time instead of switching majors and taking another year or more to finish. Or, maybe it was one of the hardest CS programs in the country, and their worst programmers are just average in the industry.
Actually, out in certain suburbs, housing prices have taken a nosedive. Check craigslist for prices in the Inland Empire (60 miles east of Los Angeles) and you'll find several houses for under $200,000 (Moreno Valley, Temecula, etc...)
Or if you use the down payment instead to pay rent five years in advance, your rent would be free! :)
This isn't true anymore. A 2BD/1BA house in Los Angeles rents for not less than $2000/month. A $400,000 mortgage at 5% is about the same. Right now you can only get a house in the cheapest nbhds for that price, but other nbhds are about to follow.
I'm the same here in Los Angeles. I had a chest CT done, and the doctor wondered if I were a smoker (I'm not). Several years of inhaling dust isn't good for you. I should start vacationing in Seattle :)
a) Of course I do, but how is that ironic? b) There are over one billion Muslims in the world. What does it matter if you personally know a couple of them? Also, what would you say if I stated that most Muslims are violent thugs because I know a few Muslims who are violent thugs?
But you don't know any, do you? And that's the point. Ignorance.
You'll do even better on file size if you convert scans to Djvu: Scan to pnm (or pgm or pbm), convert using c44 (or cjb2 for pbm), and you can skip the TeX step (use djvm to assemble pages). If the text you're scanning is black and white, you'll be stunned by the quality and file size reduction.
I don't think 300dpi is high for archival purposes. I think it's around the minimum. I can convert handwritten bitonal 8x11in 450dpi scans into 30-40 kilobyte Djvu files.
I bought a used car (private party) with 666 in the license plate. After a relative begged me to swap it, I went to the DMV (California) to change it, since it was only $8. After filling out some papers, the clerk gave me a new plate, and as I was about to leave, I realized I hadn't paid. So I told him, and he said that anyone with a 666 in their license can exchange it for free. I was pleased.
That's interesting. Here in L.A., when there are strong winds, the smog disappears and you can see the mountains. It's really beautiful. A few days later the haze is as thick as it was before. I always wondered where the fog disappears to. So I guess it rises out of view when it's windy, and falls back into place later? That seems strange.
Nice pictures! I also found this quote interesting:
:)
A lovely gin and tonic to start the flight off and the option to stretch ones legs out and work on the MacBook Air in-between looking out the window to take pictures makes every flight much more pleasant.
Is this how people usually turn out when they buy a Mac?
I know, but that's my point. People who don't find Matlab useful never post here. I never claimed that Matlab was useless.
Not sure if Windows 2008 is rock solid, but I like it because ... it's a free download, so I run it as a KVM virtual machine. Add to that Visual C++ Express Edition, and I don't have to purchase any MS software when I need it for teaching. Pretty cool.
Nice homepage, by the way!
The good thing about a tax on obesity is that the gov't would be sued into providing a proper definition. Once it's showed that it is common for healthy men of the same height and same body fat percentage to differ in weight by as much as 60 pounds, the commonly-used nonsensical definition of obesity might go away.
Quick question: How long did it take for you to lose 12 pounds doing that? And what percent of you weight was that?
Even with private health insurance, those who live unhealthy lifestyles have the net effect of increasing insurance premiums for everyone.
I think history, and expecially recent history, shows this reasoning is not true. The demand for health insurance is inelastic, just as it is for oil. Companies don't have to collude to keep prices high when they know certain helth services are not optional-- they just charge as much as possible, and will charge more in the future.