I find it ironic that your responses to my freely offered posts are so hostile; you are free to ignore those too, after all.
I think I have made it quite clear to those who actually read my posts carefully what it is that I believe.
When someone offers something for public consumption, and SourceForge is a public resource, they are inherently opening themselves to public comment. When they claim to be advancing the state of computing technology, and I believe they are not doing so, I will say so in public.
If, as a true gift, someone *personally* e-mails me with a link to a software package that I find does not suit my needs, I will of course thank them for their effort, even if I end up not using the project. If they persist and talk constantly about how much help they have been to me, I will, however, have to gently reveal the problems that I had. Honesty is sometimes the only polite response to disabuse someone of an embarrassingly mistaken notion.
When, on the other hand, a potentially useful public resource like SourceForge is made less useful by the public act of those who submit the 358th MP3 player project, I will rightfully complain.
The problem is real: how do I find the best open-source MP3 player? Must I download every one, fight to install it (later repairing any damage it does to my system), and try it out for a reasonable time? That is impractical and wasteful. Should I simply go by popularity? Well, we all know that popularity doesn't mean excellence.
To counter your supposition that I am simply a whining consumer, the problem also arises for those with talent to contribute. If I wish to improve the state of the art in MP3 playing, where do I contribute? How do I pick which player to improve, so that my effort serves the public good? The most popular? Well, then, I've simply piled on to the "good enough to be popular" solution. Inevitably, it will be a clone of a commercial product, and the *only* virtue will be its source license, or a few extra geek-friendly features. Or do I need to search through the pile for the true gem that has the kernel of a great idea, and polish that gem to perfection? I'm probably not going to be able to find that gem in the pile of junk produced by every self-indulgent geek who felt an MP3 player would be a great first project to hone their Java skills, and that first project ought to be made available to the whole universe.
What I would prefer is that the 300+ people who had inferior MP3 player projects had exercised som self-control and due diligence in researching the state of the art. Yet this responsibility of an artisan to have enough pride to not publish shoddy, incomplete work seems to be a totally absent virtue in the open source movement.
You are totally misinterpreting my motives. No where do I demand that someone do something for me. I simply comment that what they are doing is not helpful, no matter how much they believe it is.
I'm not trying to choose for anyone. However, the choices of everyone else DO affect me! There is a disturbingly finite amount of human effort, particularly of humans who can program computers.
When free software advocates think that freedom is suitable as an end *in itself*, and are willing to relax as soon as there is a libre alternative (no matter how clunky) to a given proprietary solution, then no progress is made.
Why? Because I still have essentially two choices.
1) proprietary, usable, if not ideally so 2) libre, but clunky
What about option (3) BETTER? As in *more* functional, *more* flexible, *more* empowering? Every person who gets distracted by the political issues of option (2) is one less person pursuing option (3).
I'm willing to give up libre if I can get BETTER. Political freedom(*) is a good thing (which I would *not* give up for better material circumstances) because it allows us all to experience our humanity more fully. Software freedom is only a good thing if it allows us to get better software. Not simply unencumbered by property rights, but BETTER.
(*) Don't claim I'm contradicting myself here. I'm simply making a distinction between broader human expression and the narrow, essentially technological property which is liberated by free software.
What is so hard to understand about the concept of signal-to-noise-ratio? SourceForge is unhelpful because for any given problem, there are either zero solutions you can find, or a huge number of solutions, *all* of which are probably mediocre. I would rather have one good solution to each problem.
Just to pick one example:
Text Editors
Documentation (150 projects)
Emacs (78 projects)
Integrated Development Environments (IDE) (275 projects)
Text Processing (255 projects)
Word Processors (124 projects)
275 IDE projects! Read through that list: it is almost completely a bunch of one-trick ponies. This one is good for Java. This one is good for Python. This one is good for Web pages.
Anything in there that will ever be a threat to Visual Studio? Can you really believe that this is the 21st century, yet VS is the best we can do?
Other areas are even worse: MP3 players: 358 projects! What a waste of human effort.
You seem to have latched on to my analogy without realizing you were the one who thought that "free shit" [note the quote marks, as in your words, not mine] was a good thing.
I was hardly insulting the generous people who make excellent and free tools. What I was insulting were the people who think that "freedom" is more important than "functional"---that somehow, a project that is free even though it is less than pre-alpha quality, is somehow doing me a favor by polluting cyberspace with yet more superfluous crap.
If I come over to your office and take a dump on your keyboard, are you going to thank me giving you "shit for free" (as in feces), or are you going to complain that it stinks?
A SourceForge with diamonds buried under piles of self-indulgent shit is not the same as a jeweler's display case.
I'd heard of LyX. It still uses standard LaTeX at the core, right?
I've kvetched about this before on/., but what I mean is a TeX implementation that wasn't written with pre-1980 machines in mind. TeX aspires to be a programming language, but in a lot of ways it is like assembler for a weird virtual machine. Fixed sets of 256 single-purpose registers, extremely weird syntax. That was fine when 256k of core was embarrassingly large. I'd like a text formatting language that allows me to use real programming idioms for abstraction---define real functions and subroutines, not just "macros" that work through TeX's basically fixed parsing routines.
Although I bow to Knuth's genius as a computer scientist, his software architecture shows its age. All sorts of "nifty" little tricks exist in the source code of TeX to make it practical to use on a 1980 computer, but make it hazardous to change TeX's internals: data structures and algorithms are closely tied together with little abstraction. As an example, there is basically no separation between the low-level parsing and the output of.dvi. Lots of TeX's rules are implemented in low-level Pascal with intricate flow-of-control, instead of as high-level descriptions of TeX's algorithms.
What I really want is a "TeX" that contains TeX's algorithms for page, line, and math layout, combined with abstract, flexible, and extensible data structures, using a real programming language. You couldn't do this practically in 1980, but you ought to be able to in 2002.
If you think that self-reliant freedom is the most important thing in your life, then I'm glad you get the chance to be happy. For me, that brings up echoes of North Korean self-reliant "juche" which consists of proudly and defiantly creating a dystopian (yet self-reliant!!!) un-paradise.
For me, the freedom to do everything myself is an abstract freedom that comes along with a concrete requirement that I actually *do* everything myself. As for me, my life is too f*cking short to figure out glibc dependencies, or wade through the million pre-alpha "projects" on SourceForge to find an app that does what I need at the moment. I'd rather give up on computers entirely than savor a freedom at such a high price.
When I can get a proprietary tool that *works*, I don't need the freedom to change it. The fact that I have the freedom to change a libre tool that doesn't work just means more work for me.
[As an aside, I actually do use libre tools like Emacs and TeX, and even hack at them from time to time. But man, I wish all the effort that goes into yet another skinnable MP3 player for Linux somehow could be refined and channeled to bring the power of Emacs and TeX into the 21st century, instead of the graceful aging of a fine 1970's vintage.]
1) I don't like arbitrary limits in my tools, no matter what they are called. This limit has stayed fixed for 5+ years as storage capacities have exploded. We are headed to 64-bit computing, and Excel will probably stay stuck with this 8/16 bit limit for some time. 256 is pitiful, 64k is still pretty small, don't you think?
2) Your distinction between data processing app and spreadsheet is somewhat arbitrary, and probably conditioned by the tautological definition of Excel==spreadsheet that the market has imposed.
3) In the MS view, Excel *is* the data processing app, to go along with the other VB tools used to access data bases.
4) This is just one of many further examples of where MS tools in general are only "good enough." The victory of VB is another example. [As a Lisp hacker, I have wistful dreams of programming Excel in "Visual Lisp for Applications." Cells are truly arbitrary objects, arbitrary limits are non-existent, and Excel is implemented with quality in mind, including informative documentation of the object model.]
Depends on what you mean by "great." Compared to what?
MS Word is not particularly strong as an *editor*; it is a word processor, meaning that it is supposed to produce formatting as well as text. The main weaknesses of Word are 1) the invisibility of its formatting markup, which can get bunged up in confusing ways 2) the poor control over page layout which causes all sorts of problems for long documents with many graphical elements.
As far as Excel being a "great" spreadsheet, it gets that by basically being the ONLY spreadsheet in wide use. In comparison to Lotus 1-2-3, it probably is great. Compared to what a spreadsheet COULD aspire to? It isn't so clear. Some weaknesses of Excel
1) sheet size is woefully limited. 65536 rows, 256 columns IIRC. Makes it hard to process data files with more than 65536 lines in them. 2) advanced math functions tend to be missing, buggy, or inadequate. Check out FLOOR() and CEILING(), particularly the mysterious, non-standard, and useless second argument. Complex number "support" is totally and grossly hacked. The documentation for erf is wrong.
3) Graphs are butt-ugly, and relatively hard to customize in certain ways. (Try to control the size of the axes so that multiple graphs can be overlaid on one another...)
-1) Strength: I must admit that the scriptability through VBA is pretty kick-ass, although VBA is pretty gross.
I could go on.
Re:You misunderstand completely
on
E ~ mc^2
·
· Score: 2
Evolutionary biology, as with archaeology, is an historical science. It makes claims as to what happened in the past (continuing in the present). The way to test theories in evolutionary biology is to continue studying existing organisms and fossil specemins in ways that determine their historical development.
With the tools of modern biochemistry, for instance, we can use DNA sequencing to test whether organisms that we believe to be related from previous studies actually share common DNA patterns that are consistent with common descent.
To find that the DNA sequences are incompatible or unrelated would create a difficulty that must be resolved. If it can't be resolved in the frame work of evolutionary theory, then that is disproof!
As an extreme example, if the fossil record started showing (what are currently belived to be) relatively recent forms (e.g. modern humans) in much older sediments, then that would cast serious doubt on the current picture of human descent. Given that DNA sequencing pretty convincingly links humans to other primates, and to other mammals, and the current fossil record sets pretty firm limits on the time when these various groups came into being, the possible ages of human fossils are actually pretty well constrained by current theory. That is, it isn't hard at all for a fossil discovery to disprove evolution in the case of humans. That no such fossil has yet been found is evidence, in the provisional sense of all scientific evidence.
Re:You misunderstand completely
on
E ~ mc^2
·
· Score: 5, Insightful
Did you actually read the link you included from talkorigins.org? It contains a plausible sequence of evolutionary changes that would lead to the bombardier beetle; exactly what you claim is impossible.
In any case, argument from design doesn't provide any "explanation," much less a better one. How did the designer make the beetle, and all its close genetic relatives, where none had existed before? Why the variety of mechanisms in the close relatives, instead of a single design?
Steel tariffs in America are, first and foremost, political favors meant to sway the feelings of visible, organized, often wealthy, and therefore influential groups. Steel producers are far more effective lobbyists that steel consumers, because the latter are so varied as to not be able to speak with one voice.
Look at other major tariff regimes: Do you think that textile production is "strategic"? You really think that sugar is a "strategic" resource? We in the U.S. pay much more than the world price for sugar. So that our army can count on a future sugar supply? Hardly: sugar farmers are much more organized than sugar consumers. As are owners of textile mills.
Your use of "Complexity Theory" as if it is a theory that can explain the difference between human brains and computers shows that you might have read books on it, but doesn't prove that you understood any of it.
Any useful "complexity" in a field should have to do with the concepts in a field which are actually UNDERSTOOD, and can be USED as the basis of study. For biology, for instance, the basic concepts are things like evolution, genetics, the biochemical basis for cell function, etc. For physics, the basic concepts are things like Maxwell's equations, quantum field theory, thermodynamics and statistical mechanics, Einstein's field equations, etc. What makes it complicated is that these basic equations are important in many different situations which complicated enough to defy direct solution. For instance, applying even non-relativistic quantum mechanics to just the electrons in a moderately sized atom is too complicated to solve. Instead, we need to work up from the level that can be solved (hydrogen atom, ignoring relativistic effects) and try to approximate what happens in, say, a carbon atom, giving mathematical predictions as to the atom's behavior. Similarly for plasma physics: magnetohydrodynamics is just too damn complex to try to solve directly, so you have to make approximations of various kinds. I haven't even touched the kinds of complexity facing those trying to understand superconductivity, or any other basic phenomenon in condensed matter physics.
Now, are you really saying to me that any sociologist *uses* as the basis of his *daily work*, theories that are more complicated than magnetohydrodynamics or quantum chromodynamics or even the standard model?
The fact that you can't reduce things to mathematics, even in the form of a model, means you haven't understood things rigorously enough to make *scientific* predictions. By "mathematics" I am being very broad: I mean to include, for instance, describing the base pairing structure of DNA, the DNA replication process, and protein synthesis as "mathematical" models. They have a precise logical structure, clearly related to the underlying physical constituents, and allow logical reasoning to develop testable hypotheses. As opposed to non-mathematical models such as the Hegelian dialectic underlying Marxist theory. Sure, you can describe a society in Marxist terms, but the structure isn't precise enough to make any testable predictions, as opposed to political arguments. That's why it isn't a "soft" science, it just isn't a science at all.
I'm sure you think Critical Theory is the most complicated of all, right? It's so damn "complicated" that not a single *useful* development has been produced by it. Don't confuse "complexity" with "mental masturbation."
Great theory, except the Dept. of Defense hardly cares about steel these days. Much more about software and remote sensing. Plus, we don't need as much untrained labor to feed the infantry either.
"Sociology" a complex discipline? Physics as simple? That, frankly, is a load of crap.
I think when you say "complex" you mean "devoid of rigor" or "full of whatever trendy crap someone felt like spouting." Hell, even using the word "discipline" is a stretch for sociology.
The difference between "hard" and "soft" sciences has nothing to do with sociology, which doesn't qualify as either. If you don't understand that, you need to review your definition of "science."
To elaborate on the AC who hasn't been modded up yet, what makes a difference is "antenna gain." The gain/directivity of an antenna is proportional to the antenna effective area divided by the wavelength squared. By shrinking the wavelength, you greatly increase your ability to "focus" the beam power into a narrow angle. It's hard to make optical dishes as large as radio dishes or arrays, but we are talking about a factor of something like 10^10 in your favor.
To exploit this in a communications link, you need better pointing accuracy for shorter wavelengths. This isn't really a problem for interstellar communication: you aim at the very visible star, and allow the beam to broaden to a few AU in diameter at the target, to get nearby planets. In radio, you would waste a bunch of power far from the star, where there won't be planets to receive it.
Also, you can pulse optical sources to create very high peak powers for short durations. (Bright flashes instead of dim and steady).
For further discussion, Horowitz has written this paper.
well, the density of the interplanetary or interstellar medium is somewhat higher than the density of the universe; Pioneer is not in intergalactic space. A better estimate would be about 1 atom/cm^3, (about 6*10^6 amu/m^3 at 1 AU from the sun, at 80 AU is is about 10^-4 of this, and guessing that this near the heliopause the solar wind contribution is roughly equal to the interstellar background) or about 10^-23 kg/m^3.
But your point is still accurate, as long as 10^7 is considerable.
The U.S. had nothing to do with Fermi's building of the first fission reactor? That's interesting, I thought it was part of the U.S. project to develop the atomic bomb.
Sure, Fermi won his Nobel Prize before he came to the U.S. But before he came to America, Fermi's accomplishments were only in basic science.
Incidentally, these world-class scientists laid the foundation for the post-war excellence of U.S. scientific institutions. Germany has hardly the scientific impact these days (even 50 years later) that it did before the war. Europe's loss was America's gain.
In principle, the goal of Allied bombing was to target economic and military production, not necessarily to kill Germans. Bombers had a chance to actually see what they were trying to hit. The V2 could only be roughly targeted.
According to Wikipedia, more than 1300 V2 rockets were fired at London. Iraq launched something like 40 Scuds toward each of Israel and Saudia Arabia in the Gulf war.
The Scud missile is basically a modified V2 in design, gaining additional range by reducing accuracy and warhead weight.
I am a physicist by training, in a field which didn't go over to preprint servers while I was in it. Preprint servers are good ways for groups to post their current results. By what strategy do you read preprint servers, though? From what I saw of my colleagues, you had to apply a skepticism rating that was generally based on reputation, and the preprints gave you a rough idea of what they thought they had achieved or had in progress. Unfortunately, there was no way to know if they eventually found some big problem in their analysis, unless you talked to them at conferences, called them up, or noticed that the result never ended up published under peer review, all of which you could do without arXiv. ArXiv.org doesn't magically break the rule that 99% of everything is crap. It just increases volume without increasing signal-to-noise.
The slashdot reference in my post was a response to someone suggesting a slashdot moderation system to move preprint servers toward peer review, not a response to the current state of preprint servers.
Technically, the V2 was a great advance. However, as a weapon, it was more psychological than military. Biological warheads are a major technical problem of their own, and the guidance was crude. Plus, by Sept. 1944, the first date for a successful V2 attack, the outcome of the war in Europe was no longer in doubt. As an aside, the stats I've seen show that roughly a dozen civilians were killed or injured on average by each V2 that hit London.
I generally agree with your points. Von Braun isn't the greatest example of human achievement, but it was the choice of the original poster, not mine. But still, isn't it amazing that something based on missile throw weights led to something as sublime as men on the moon?
I thought it was pretty clear from my post that I'm in favor of liberal immigration laws, and against the kind of oppressive political scene that the FBI harrassment of Einstein represents.
In terms of national interests, it does matter to me where there is a prosperous, free, and open society. If the U.S. starts to be less of any of these, I'm going to work to change that, because the U.S. is where I happen to be, my family and friends live here, and I would rather not uproot and move somewhere else if I can avoid it. No one wants to be a refugee, or an emigrant, if they could avoid it.
Slashdot as a model? You must be out of your mind. Slashdot has almost uniformly worked against the possibility of informative, rigorously reasoned discourse.
I agree that scientific journals are in trouble, and the means of their survival are not at all clear. But I am skeptical that arXiv is a model that will be successful. It is instead corrosive to methodical, rigorous, careful investigation, and promotes flashy, hasty, and superficial discussion. (Much like slashdot?) "Publish early, publish often" is the arXiv game.
Re:Check out arXiv.org
on
Who Owns Science?
·
· Score: 4, Interesting
"greedy journal publishers" is pure flamebait. What is this, an argument about record labels?
The problem with arXiv is that much of the stuff on there would not pass peer-review, and some of it never gets revised to pass muster. By the time the author gets around to publishing it in a peer-reviewed journal, the on-line preprints have moved on, so the topic is no longer considered worth the effort of publication.
The end result is that all the readers of preprint servers have to do their own peer review, which is incredibly wasteful of effort.
Journal publishers are *not* making any kind of outrageous profits. Instead, they are defraying the substantial costs they incur in managing the editorial process that keeps scientific journals from becoming cesspools of "we publish anything!!!"
Slashdot Mirror
I find it ironic that your responses to my freely offered posts are so hostile; you are free to ignore those too, after all.
I think I have made it quite clear to those who actually read my posts carefully what it is that I believe.
When someone offers something for public consumption, and SourceForge is a public resource, they are inherently opening themselves to public comment. When they claim to be advancing the state of computing technology, and I believe they are not doing so, I will say so in public.
If, as a true gift, someone *personally* e-mails me with a link to a software package that I find does not suit my needs, I will of course thank them for their effort, even if I end up not using the project. If they persist and talk constantly about how much help they have been to me, I will, however, have to gently reveal the problems that I had. Honesty is sometimes the only polite response to disabuse someone of an embarrassingly mistaken notion.
When, on the other hand, a potentially useful public resource like SourceForge is made less useful by the public act of those who submit the 358th MP3 player project, I will rightfully complain.
The problem is real: how do I find the best open-source MP3 player? Must I download every one, fight to install it (later repairing any damage it does to my system), and try it out for a reasonable time? That is impractical and wasteful. Should I simply go by popularity? Well, we all know that popularity doesn't mean excellence.
To counter your supposition that I am simply a whining consumer, the problem also arises for those with talent to contribute. If I wish to improve the state of the art in MP3 playing, where do I contribute? How do I pick which player to improve, so that my effort serves the public good? The most popular? Well, then, I've simply piled on to the "good enough to be popular" solution. Inevitably, it will be a clone of a commercial product, and the *only* virtue will be its source license, or a few extra geek-friendly features. Or do I need to search through the pile for the true gem that has the kernel of a great idea, and polish that gem to perfection? I'm probably not going to be able to find that gem in the pile of junk produced by every self-indulgent geek who felt an MP3 player would be a great first project to hone their Java skills, and that first project ought to be made available to the whole universe.
What I would prefer is that the 300+ people who had inferior MP3 player projects had exercised som self-control and due diligence in researching the state of the art. Yet this responsibility of an artisan to have enough pride to not publish shoddy, incomplete work seems to be a totally absent virtue in the open source movement.
You are totally misinterpreting my motives. No where do I demand that someone do something for me. I simply comment that what they are doing is not helpful, no matter how much they believe it is.
I'm not trying to choose for anyone. However, the choices of everyone else DO affect me! There is a disturbingly finite amount of human effort, particularly of humans who can program computers.
When free software advocates think that freedom is suitable as an end *in itself*, and are willing to relax as soon as there is a libre alternative (no matter how clunky) to a given proprietary solution, then no progress is made.
Why? Because I still have essentially two choices.
1) proprietary, usable, if not ideally so
2) libre, but clunky
What about option (3) BETTER? As in *more* functional, *more* flexible, *more* empowering?
Every person who gets distracted by the political issues of option (2) is one less person pursuing option (3).
I'm willing to give up libre if I can get BETTER. Political freedom(*) is a good thing (which I would *not* give up for better material circumstances) because it allows us all to experience our humanity more fully. Software freedom is only a good thing if it allows us to get better software. Not simply unencumbered by property rights, but BETTER.
(*) Don't claim I'm contradicting myself here. I'm simply making a distinction between broader human expression and the narrow, essentially technological property which is liberated by free software.
What is so hard to understand about the concept of signal-to-noise-ratio? SourceForge is unhelpful because for any given problem, there are either zero solutions you can find, or a huge number of solutions, *all* of which are probably mediocre. I would rather have one good solution to each problem.
Just to pick one example:
Text Editors
Documentation (150 projects)
Emacs (78 projects)
Integrated Development Environments (IDE) (275 projects)
Text Processing (255 projects)
Word Processors (124 projects)
275 IDE projects! Read through that list: it is almost completely a bunch of one-trick ponies. This one is good for Java. This one is good for Python. This one is good for Web pages.
Anything in there that will ever be a threat to Visual Studio? Can you really believe that this is the 21st century, yet VS is the best we can do?
Other areas are even worse: MP3 players: 358 projects! What a waste of human effort.
You seem to have latched on to my analogy without realizing you were the one who thought that "free shit" [note the quote marks, as in your words, not mine] was a good thing.
I was hardly insulting the generous people who make excellent and free tools. What I was insulting were the people who think that "freedom" is more important than "functional"---that somehow, a project that is free even though it is less than pre-alpha quality, is somehow doing me a favor by polluting cyberspace with yet more superfluous crap.
If I come over to your office and take a dump on your keyboard, are you going to thank me giving you "shit for free" (as in feces), or are you going to complain that it stinks?
A SourceForge with diamonds buried under piles of self-indulgent shit is not the same as a jeweler's display case.
I'd heard of LyX. It still uses standard LaTeX at the core, right?
/., but what I mean is a TeX implementation that wasn't written with pre-1980 machines in mind. TeX aspires to be a programming language, but in a lot of ways it is like assembler for a weird virtual machine. Fixed sets of 256 single-purpose registers, extremely weird syntax. That was fine when 256k of core was embarrassingly large. I'd like a text formatting language that allows me to use real programming idioms for abstraction---define real functions and subroutines, not just "macros" that work through TeX's basically fixed parsing routines.
.dvi. Lots of TeX's rules are implemented in low-level Pascal with intricate flow-of-control, instead of as high-level descriptions of TeX's algorithms.
I've kvetched about this before on
Although I bow to Knuth's genius as a computer scientist, his software architecture shows its age. All sorts of "nifty" little tricks exist in the source code of TeX to make it practical to use on a 1980 computer, but make it hazardous to change TeX's internals: data structures and algorithms are closely tied together with little abstraction. As an example, there is basically no separation between the low-level parsing and the output of
What I really want is a "TeX" that contains TeX's algorithms for page, line, and math layout, combined with abstract, flexible, and extensible data structures, using a real programming language. You couldn't do this practically in 1980, but you ought to be able to in 2002.
If you think that self-reliant freedom is the most important thing in your life, then I'm glad you get the chance to be happy. For me, that brings up echoes of North Korean self-reliant "juche" which consists of proudly and defiantly creating a dystopian (yet self-reliant!!!) un-paradise.
For me, the freedom to do everything myself is an abstract freedom that comes along with a concrete requirement that I actually *do* everything myself. As for me, my life is too f*cking short to figure out glibc dependencies, or wade through the million pre-alpha "projects" on SourceForge to find an app that does what I need at the moment. I'd rather give up on computers entirely than savor a freedom at such a high price.
When I can get a proprietary tool that *works*, I don't need the freedom to change it. The fact that I have the freedom to change a libre tool that doesn't work just means more work for me.
[As an aside, I actually do use libre tools like Emacs and TeX, and even hack at them from time to time. But man, I wish all the effort that goes into yet another skinnable MP3 player for Linux somehow could be refined and channeled to bring the power of Emacs and TeX into the 21st century, instead of the graceful aging of a fine 1970's vintage.]
I can answer your comment in several ways
1) I don't like arbitrary limits in my tools, no matter what they are called. This limit has stayed fixed for 5+ years as storage capacities have exploded. We are headed to 64-bit computing, and Excel will probably stay stuck with this 8/16 bit limit for some time. 256 is pitiful, 64k is still pretty small, don't you think?
2) Your distinction between data processing app and spreadsheet is somewhat arbitrary, and probably conditioned by the tautological definition of Excel==spreadsheet that the market has imposed.
3) In the MS view, Excel *is* the data processing app, to go along with the other VB tools used to access data bases.
4) This is just one of many further examples of where MS tools in general are only "good enough." The victory of VB is another example. [As a Lisp hacker, I have wistful dreams of programming Excel in "Visual Lisp for Applications." Cells are truly arbitrary objects, arbitrary limits are non-existent, and Excel is implemented with quality in mind, including informative documentation of the object model.]
5) I've cursed this limit several times.
Depends on what you mean by "great." Compared to what?
MS Word is not particularly strong as an *editor*; it is a word processor, meaning that it is supposed to produce formatting as well as text.
The main weaknesses of Word are 1) the invisibility of its formatting markup, which can get bunged up in confusing ways 2) the poor control over page layout which causes all sorts of problems for long documents with many graphical elements.
As far as Excel being a "great" spreadsheet, it gets that by basically being the ONLY spreadsheet in wide use. In comparison to Lotus 1-2-3, it probably is great. Compared to what a spreadsheet COULD aspire to? It isn't so clear. Some weaknesses of Excel
1) sheet size is woefully limited. 65536 rows, 256 columns IIRC. Makes it hard to process data files with more than 65536 lines in them.
2) advanced math functions tend to be missing, buggy, or inadequate. Check out FLOOR() and CEILING(), particularly the mysterious, non-standard, and useless second argument. Complex number "support" is totally and grossly hacked.
The documentation for erf is wrong.
3) Graphs are butt-ugly, and relatively hard to customize in certain ways. (Try to control the size of the axes so that multiple graphs can be overlaid on one another...)
-1) Strength: I must admit that the scriptability through VBA is pretty kick-ass, although VBA is pretty gross.
I could go on.
Evolutionary biology, as with archaeology, is an historical science. It makes claims as to what happened in the past (continuing in the present). The way to test theories in evolutionary biology is to continue studying existing organisms and fossil specemins in ways that determine their historical development.
With the tools of modern biochemistry, for instance, we can use DNA sequencing to test whether organisms that we believe to be related from previous studies actually share common DNA patterns that are consistent with common descent.
To find that the DNA sequences are incompatible or unrelated would create a difficulty that must be resolved. If it can't be resolved in the frame work of evolutionary theory, then that is disproof!
As an extreme example, if the fossil record started showing (what are currently belived to be) relatively recent forms (e.g. modern humans) in much older sediments, then that would cast serious doubt on the current picture of human descent. Given that DNA sequencing pretty convincingly links humans to other primates, and to other mammals, and the current fossil record sets pretty firm limits on the time when these various groups came into being, the possible ages of human fossils are actually pretty well constrained by current theory. That is, it isn't hard at all for a fossil discovery to disprove evolution in the case of humans. That no such fossil has yet been found is evidence, in the provisional sense of all scientific evidence.
Did you actually read the link you included from talkorigins.org? It contains a plausible sequence of evolutionary changes that would lead to the bombardier beetle; exactly what you claim is impossible.
In any case, argument from design doesn't provide any "explanation," much less a better one. How did the designer make the beetle, and all its close genetic relatives, where none had existed before? Why the variety of mechanisms in the close relatives, instead of a single design?
Steel tariffs in America are, first and foremost, political favors meant to sway the feelings of visible, organized, often wealthy, and therefore influential groups. Steel producers are far more effective lobbyists that steel consumers, because the latter are so varied as to not be able to speak with one voice.
Look at other major tariff regimes: Do you think that textile production is "strategic"? You really think that sugar is a "strategic" resource? We in the U.S. pay much more than the world price for sugar. So that our army can count on a future sugar supply? Hardly: sugar farmers are much more organized than sugar consumers. As are owners of textile mills.
Your use of "Complexity Theory" as if it is a theory that can explain the difference between human brains and computers shows that you might have read books on it, but doesn't prove that you understood any of it.
Any useful "complexity" in a field should have to do with the concepts in a field which are actually UNDERSTOOD, and can be USED as the basis of study. For biology, for instance, the basic concepts are things like evolution, genetics, the biochemical basis for cell function, etc. For physics, the basic concepts are things like Maxwell's equations, quantum field theory, thermodynamics and statistical mechanics, Einstein's field equations, etc. What makes it complicated is that these basic equations are important in many different situations which complicated enough to defy direct solution. For instance, applying even non-relativistic quantum mechanics to just the electrons in a moderately sized atom is too complicated to solve. Instead, we need to work up from the level that can be solved (hydrogen atom, ignoring relativistic effects) and try to approximate what happens in, say, a carbon atom, giving mathematical predictions as to the atom's behavior. Similarly for plasma physics: magnetohydrodynamics is just too damn complex to try to solve directly, so you have to make approximations of various kinds. I haven't even touched the kinds of complexity facing those trying to understand superconductivity, or any other basic phenomenon in condensed matter physics.
Now, are you really saying to me that any sociologist *uses* as the basis of his *daily work*, theories that are more complicated than magnetohydrodynamics or quantum chromodynamics or even the standard model?
The fact that you can't reduce things to mathematics, even in the form of a model, means you haven't understood things rigorously enough to make *scientific* predictions. By "mathematics" I am being very broad: I mean to include, for instance, describing the base pairing structure of DNA, the DNA replication process, and protein synthesis as "mathematical" models. They have a precise logical structure, clearly related to the underlying physical constituents, and allow logical reasoning to develop testable hypotheses. As opposed to non-mathematical models such as the Hegelian dialectic underlying Marxist theory. Sure, you can describe a society in Marxist terms, but the structure isn't precise enough to make any testable predictions, as opposed to political arguments. That's why it isn't a "soft" science, it just isn't a science at all.
I'm sure you think Critical Theory is the most complicated of all, right? It's so damn "complicated" that not a single *useful* development has been produced by it. Don't confuse "complexity" with "mental masturbation."
Great theory, except the Dept. of Defense hardly cares about steel these days. Much more about software and remote sensing. Plus, we don't need as much untrained labor to feed the infantry either.
"Sociology" a complex discipline? Physics as simple? That, frankly, is a load of crap.
I think when you say "complex" you mean "devoid of rigor" or "full of whatever trendy crap someone felt like spouting." Hell, even using the word "discipline" is a stretch for sociology.
The difference between "hard" and "soft" sciences has nothing to do with sociology, which doesn't qualify as either. If you don't understand that, you need to review your definition of "science."
To elaborate on the AC who hasn't been modded up yet, what makes a difference is "antenna gain." The gain/directivity of an antenna is proportional to the antenna effective area divided by the wavelength squared. By shrinking the wavelength, you greatly increase your ability to "focus" the beam power into a narrow angle. It's hard to make optical dishes as large as radio dishes or arrays, but we are talking about a factor of something like 10^10 in your favor.
To exploit this in a communications link, you need better pointing accuracy for shorter wavelengths. This isn't really a problem for interstellar communication: you aim at the very visible star, and allow the beam to broaden to a few AU in diameter at the target, to get nearby planets. In radio, you would waste a bunch of power far from the star, where there won't be planets to receive it.
Also, you can pulse optical sources to create very high peak powers for short durations. (Bright flashes instead of dim and steady).
For further discussion, Horowitz has written this paper.
well, the density of the interplanetary or interstellar medium is somewhat higher than the density of the universe; Pioneer is not in intergalactic space. A better estimate would be about 1 atom/cm^3, (about 6*10^6 amu/m^3 at 1 AU from the sun, at 80 AU is is about 10^-4 of this, and guessing that this near the heliopause the solar wind contribution is roughly equal to the interstellar background) or about 10^-23 kg/m^3.
But your point is still accurate, as long as 10^7 is considerable.
The U.S. had nothing to do with Fermi's building of the first fission reactor? That's interesting, I thought it was part of the U.S. project to develop the atomic bomb.
Sure, Fermi won his Nobel Prize before he came to the U.S. But before he came to America, Fermi's accomplishments were only in basic science.
Incidentally, these world-class scientists laid the foundation for the post-war excellence of U.S. scientific institutions. Germany has hardly the scientific impact these days (even 50 years later) that it did before the war. Europe's loss was America's gain.
In principle, the goal of Allied bombing was to target economic and military production, not necessarily to kill Germans. Bombers had a chance to actually see what they were trying to hit. The V2 could only be roughly targeted.
According to Wikipedia, more than 1300 V2 rockets were fired at London. Iraq launched something like 40 Scuds toward each of Israel and Saudia Arabia in the Gulf war.
The Scud missile is basically a modified V2 in design, gaining additional range by reducing accuracy and warhead weight.
I am a physicist by training, in a field which didn't go over to preprint servers while I was in it. Preprint servers are good ways for groups to post their current results. By what strategy do you read preprint servers, though? From what I saw of my colleagues, you had to apply a skepticism rating that was generally based on reputation, and the preprints gave you a rough idea of what they thought they had achieved or had in progress. Unfortunately, there was no way to know if they eventually found some big problem in their analysis, unless you talked to them at conferences, called them up, or noticed that the result never ended up published under peer review, all of which you could do without arXiv. ArXiv.org doesn't magically break the rule that 99% of everything is crap. It just increases volume without increasing signal-to-noise.
The slashdot reference in my post was a response to someone suggesting a slashdot moderation system to move preprint servers toward peer review, not a response to the current state of preprint servers.
Technically, the V2 was a great advance. However, as a weapon, it was more psychological than military. Biological warheads are a major technical problem of their own, and the guidance was crude. Plus, by Sept. 1944, the first date for a successful V2 attack, the outcome of the war in Europe was no longer in doubt. As an aside, the stats I've seen show that roughly a dozen civilians were killed or injured on average by each V2 that hit London.
I generally agree with your points. Von Braun isn't the greatest example of human achievement, but it was the choice of the original poster, not mine. But still, isn't it amazing that something based on missile throw weights led to something as sublime as men on the moon?
I thought it was pretty clear from my post that I'm in favor of liberal immigration laws, and against the kind of oppressive political scene that the FBI harrassment of Einstein represents.
In terms of national interests, it does matter to me where there is a prosperous, free, and open society. If the U.S. starts to be less of any of these, I'm going to work to change that, because the U.S. is where I happen to be, my family and friends live here, and I would rather not uproot and move somewhere else if I can avoid it. No one wants to be a refugee, or an emigrant, if they could avoid it.
Slashdot as a model? You must be out of your mind. Slashdot has almost uniformly worked against the possibility of informative, rigorously reasoned discourse.
I agree that scientific journals are in trouble, and the means of their survival are not at all clear. But I am skeptical that arXiv is a model that will be successful. It is instead corrosive to methodical, rigorous, careful investigation, and promotes flashy, hasty, and superficial discussion. (Much like slashdot?) "Publish early, publish often" is the arXiv game.
"greedy journal publishers" is pure flamebait. What is this, an argument about record labels?
The problem with arXiv is that much of the stuff on there would not pass peer-review, and some of it never gets revised to pass muster. By the time the author gets around to publishing it in a peer-reviewed journal, the on-line preprints have moved on, so the topic is no longer considered worth the effort of publication.
The end result is that all the readers of preprint servers have to do their own peer review, which is incredibly wasteful of effort.
Journal publishers are *not* making any kind of outrageous profits. Instead, they are defraying the substantial costs they incur in managing the editorial process that keeps scientific journals from becoming cesspools of "we publish anything!!!"