Mathematica wasn't so much ported to Nextstep as it was one of the original platforms on which Mathematica was released (Macintosh System 6 being one of the others). In fact, Mathematica was bundled with each Nextstation.
It was probably pretty easy to bring Mathematica to OS X because they already had an Openstep version 3 (Cocoa is really just a newer version of Openstep). I will be using the OS X version of Mathematica soon, once our site license kicks in in June.
I just bought matlab for linux ($1900!) and I didn't find the licensing to be difficult. It took me about 5 minutes to set up. Granted, there is no GUI for the license install/configuration step, but that really shouldn't be an obstacle. I will grant that for $1900, you'd think they could come up with a nice interface for _all_ parts of the install process.
I agree about octave; I've been using it for a while and recently started using matlab again, and BOY is there a big difference (especially when it comes to graphics, and special things like sparse matrices).
..that I can get the OS X version for free. I just spent $1900 on an individual commercial linux license for my home-based scientific consulting biz, but I've got a dual G4/533 in addition to my dual P4/1.7 at home, and a mac in the office at school as well. The individual licenses actually allow you to install on as many machines as you want, as long as you agree to be the sole user and use only one copy at a time - but it may be platform-restricted.
All software has bugs and needs to be modified/fixed when it fails (or least fails to meet the users' needs), so therefore software always costs money, even after its already "done."
The primary advantange of Fortran is its incredible compilers. Fortran compilers can handle all sorts of optimizations automatically that can't be done by most C or C++ compilers.
On the other hand, managing complex data structures is a beast with Fortran (who wants function calls with 75 arguments? I've seen it done).
The best solution is to use OOP languages (C++ being the prime example in scientific computing) for handling the data structures, dynamic memor management, I/O, UI, etc. and use Fortran for all the inner loops where 90%+ of the execution time is spent.
As ESR has said in his Python... most programs nowadays don't even need compiled speed. 90% of the stuff joe user does is I/O bound, not CPU bound.
Scientific code is a special case, I suppose...I'm talking about code that takes days or weeks to run on a small cluster. The speed of C++ code depends more heavily on the compiler than the speed of C code does, but in my experience you can write good fast scientific code in C++, with a substantial savings in development and debugging time.
The problem with dynamic typing is that its not as fast as static typing. I'm writing a bunch of numerical code (for scientific research and consulting), and I use C (or C++ sometimes) but never Objective-C...when you need high performance, you want everything static if possible.
Of course, the happy middle ground is to write all the computational code in C/C++/Fortran and then make a nice GUI with Cocoa/Objective-C.
Here's a little tidbit of info that demonstrates how completely filled with misinformation your post is:
You complain bitterly how "their hardware is cruddier [sic] each year," giving the Blue and White G3 as an example. (I won't even comment on your assertions about the B+W G3's quality).
The blue and white G3 was discontinued more than two years ago!
The criticism "stinks of creationism" is perfectly valid. Creationism has a long history of using scientific-like (but unsound) reasoning to argue religious points of view. In that and other regards creationism is unlike the rest of science: instead of coming up with theories to explain the evidence, creationists already have the theory worked out and try to show how _all_ the evidence supports their theory and no others, i.e. its not really science at all.
But, the big-bang theory was indeed arrived at as a logical conclusion of theories that were constructed to explain already-observed phenonema (like the Hubble effect). So even though it looks like creationism, its not. It still could be wrong, of course.
Here's an especially tricky problem: how to transfer files larger than 2GB between Mac OS X machines and linux boxes that are thousands of miles apart, without using a tape drive.
So far, my only solution is to use tar with firewire drives: that is to say, on the source machine you tar the files directly to the device file, and then on the destination machine you untar the files again directly from the device file --- no filesystems required, just raw I/O of bytes!
1. Fortran as a language is much more optimizable by the compiler than C or especially C++, and as a result relatively naive Fortran code runs as fast as hand-optimized C code.
2. Experimental physicists aren't in the business of writing software, so they'll continue to use whatever tools they already have to do what they need to do, and will write new software in the quickest, easiest way possible (i.e. in Fortran since all the important scientific codes are written in Fortran).
Ummm, actually the reason Apple chose BSD is because Nextstep is a BSD-like unix, and Mac OS X is really just Nextstep at its core.
Re:Scientifically explained Artificial Intelligenc
on
Son of HAL For Sale
·
· Score: 1
"physical mathematical limitations"...which is it, physical or mathematical?
"fuzzy logic mathematical heterogeneric virtual input/output systems"...if slashdotters couldn't tell you were full of shit by now, they must be smoking kerosene-laced crack.
"equations involved are enormous though"...yeah, we mathematicians have a hard time with big equations, if it can't fit on one page it's just TOO BIG.
Tell me, who is your dealer...I'm sure lots of people would like to get as high as you.
As a Ph.D. applied mathematician whose research specialty is wavelet compression methods, I have some comments.
The website clearly indicates that this software is designed for images that come from scanning papers documents that contain both text and graphics. The algorithm is supposed to recognize the text and store it as a separate layer (but still as an image) from the rest of the image. Furthermore, the image can be transmitted in such as way that the text and significant features of the image are transmitted first, followed by the bits representing less important features later - a technique known as progressive image transmission (among other names).
I certainly believe that they use wavelets to do this. In fact, the hot wavelet method of the 90's, EZW (Embedded-Zero tree Wavelet) compression allows for exactly that: to compress an image in such a way that the more significant bits come first, followed by the less significant bits in order of significance. Picking out text that is layered on top of a background image is relatively easy with wavelets: just pick out the really large coefficients in the wavelet expansion, since those most likely correspond to parts of the image where large jump discontinuities occur. This can all be done automatically, of course.
Clustering IS STILL brute-force computing. All clustering does is make a more powerful computer; but it can't do any computations that a serial computer can't do. The only advantage clustering offers is more computing power, which is the same advantage that you get by using a faster serial computer.
This stuff sounds a lot like the science fiction of the forties and fifties..."By 1980, people will travel to their offices in automated helicopters! People will fly to the moon and mars on a regular basis!" Many science fiction novels of the time portrayed extravagant space-travel, and yet had humans doing the navigational computations on board these spacecraft, completely missing the coming dominance of electronic computers in computations. The popular imagination of future technology was simply an extrapolation of the current technology - "Fast airplanes, bigger rockets, atomic power!" - when in reality the technology of the future has turned out to be things that people could not have imagined at the time (the world-wide-web, biotechnology, fax machines, massive computational power...)
In the same vein, the breakthrough technology of the next 100 years will most likely consist of things that haven't even occured to us yet -- not simply faster and faster computers, smaller and smaller robots, better and better bioengineering, but rather something else entirely.
In grad school I took a unix systems administration class from Evi Nemeth (one of the authors of the book). This was 5 years ago, so we used the Red version of the book. While the book was indeed very useful, I found and continue to find that I have to look at man pages or other info on the web if I really want to know the details of how to do something with a system.
Also, I think she didn't like me, because unlike the undergrads in the class, I was willing to challenge her on things (like, "Why are you always late to class, professor?")
The point of the neural net these people have set up is not to have a neural net that recognizes words well; nor is it to have a neural net to recognize words in noisy environments from different speakers. Rather, the point of the research is to study how biological neurons can process sensory information over time to produce a sensation of a "moment." Recognizing the word "one" is an example of such temporal sensory processing, and a relatively easy one to test, so that's why they chose that task for their research project.
Mac OS X is isn't totally unrelated to MkLinux. MkLinux is based on the Mach microkernel, and Apple leveraged their development on the MkLinux project to improve the Mach kernel they are using in Mac OS X.
That is an interesting observation that you made about education level correlating with the use of "HA HA HA." As it happens, I have a Ph.D., so I guess I must be an outlier in your study.
Collage carrier! HA HA HA !! That is one of the funniest mis-spellings I've ever seen on slashdot. You may be a college student, but you sure aren't very well educated.
(Collage carrier! That sounds like a special rack designed for carrying miniature flower bouquets).
Slashdot Mirror
Mathematica wasn't so much ported to Nextstep as it was one of the original platforms on which Mathematica was released (Macintosh System 6 being one of the others). In fact, Mathematica was bundled with each Nextstation.
It was probably pretty easy to bring Mathematica to OS X because they already had an Openstep version 3 (Cocoa is really just a newer version of Openstep). I will be using the OS X version of Mathematica soon, once our site license kicks in in June.
I just bought matlab for linux ($1900!) and I didn't find the licensing to be difficult. It took me about 5 minutes to set up. Granted, there is no GUI for the license install/configuration step, but that really shouldn't be an obstacle. I will grant that for $1900, you'd think they could come up with a nice interface for _all_ parts of the install process.
I agree about octave; I've been using it for a while and recently started using matlab again, and BOY is there a big difference (especially when it comes to graphics, and special things like sparse matrices).
..that I can get the OS X version for free. I just spent $1900 on an individual commercial linux license for my home-based scientific consulting biz, but I've got a dual G4/533 in addition to my dual P4/1.7 at home, and a mac in the office at school as well. The individual licenses actually allow you to install on as many machines as you want, as long as you agree to be the sole user and use only one copy at a time - but it may be platform-restricted.
I know! Name them after characters in the Lord of the Rings. All your hax0r friends will think you are cool, hip, and original.
All software has bugs and needs to be modified/fixed when it fails (or least fails to meet the users' needs), so therefore software always costs money, even after its already "done."
Dude, you should get a Mac and run OS X...you'd never have to reboot into a different OS to a particular app, you could run it all under one OS.
The primary advantange of Fortran is its incredible compilers. Fortran compilers can handle all sorts of optimizations automatically that can't be done by most C or C++ compilers.
On the other hand, managing complex data structures is a beast with Fortran (who wants function calls with 75 arguments? I've seen it done).
The best solution is to use OOP languages (C++ being the prime example in scientific computing) for handling the data structures, dynamic memor management, I/O, UI, etc. and use Fortran for all the inner loops where 90%+ of the execution time is spent.
As ESR has said in his Python... most programs nowadays don't even need compiled speed. 90% of the stuff joe user does is I/O bound, not CPU bound.
Scientific code is a special case, I suppose...I'm talking about code that takes days or weeks to run on a small cluster. The speed of C++ code depends more heavily on the compiler than the speed of C code does, but in my experience you can write good fast scientific code in C++, with a substantial savings in development and debugging time.
The problem with dynamic typing is that its not as fast as static typing. I'm writing a bunch of numerical code (for scientific research and consulting), and I use C (or C++ sometimes) but never Objective-C...when you need high performance, you want everything static if possible.
Of course, the happy middle ground is to write all the computational code in C/C++/Fortran and then make a nice GUI with Cocoa/Objective-C.
Here's a little tidbit of info that demonstrates how completely filled with misinformation your post is:
You complain bitterly how "their hardware is cruddier [sic] each year," giving the Blue and White G3 as an example. (I won't even comment on your assertions about the B+W G3's quality).
The blue and white G3 was discontinued more than two years ago!
I could go on (and on and on...), but I won't.
The criticism "stinks of creationism" is perfectly valid. Creationism has a long history of using scientific-like (but unsound) reasoning to argue religious points of view. In that and other regards creationism is unlike the rest of science: instead of coming up with theories to explain the evidence, creationists already have the theory worked out and try to show how _all_ the evidence supports their theory and no others, i.e. its not really science at all.
But, the big-bang theory was indeed arrived at as a logical conclusion of theories that were constructed to explain already-observed phenonema (like the Hubble effect). So even though it looks like creationism, its not. It still could be wrong, of course.
NFS would work, as long as you don't mind waiting 80+ hours to transfer a 40 GB file over a T1 line...Fedex overnight is a lot faster than that.
Here's an especially tricky problem: how to transfer files larger than 2GB between Mac OS X machines and linux boxes that are thousands of miles apart, without using a tape drive.
So far, my only solution is to use tar with firewire drives: that is to say, on the source machine you tar the files directly to the device file, and then on the destination machine you untar the files again directly from the device file --- no filesystems required, just raw I/O of bytes!
Who is this "Mac" guy you are talking about...oh wait, do you mean Apple, Inc., the company that manufactures the Macintosh?
1. Fortran as a language is much more optimizable by the compiler than C or especially C++, and as a result relatively naive Fortran code runs as fast as hand-optimized C code.
2. Experimental physicists aren't in the business of writing software, so they'll continue to use whatever tools they already have to do what they need to do, and will write new software in the quickest, easiest way possible (i.e. in Fortran since all the important scientific codes are written in Fortran).
Ummm, actually the reason Apple chose BSD is because Nextstep is a BSD-like unix, and Mac OS X is really just Nextstep at its core.
"fuzzy logic mathematical heterogeneric virtual input/output systems"...if slashdotters couldn't tell you were full of shit by now, they must be smoking kerosene-laced crack.
"equations involved are enormous though"...yeah, we mathematicians have a hard time with big equations, if it can't fit on one page it's just TOO BIG.
Tell me, who is your dealer...I'm sure lots of people would like to get as high as you.
The website clearly indicates that this software is designed for images that come from scanning papers documents that contain both text and graphics. The algorithm is supposed to recognize the text and store it as a separate layer (but still as an image) from the rest of the image. Furthermore, the image can be transmitted in such as way that the text and significant features of the image are transmitted first, followed by the bits representing less important features later - a technique known as progressive image transmission (among other names).
I certainly believe that they use wavelets to do this. In fact, the hot wavelet method of the 90's, EZW (Embedded-Zero tree Wavelet) compression allows for exactly that: to compress an image in such a way that the more significant bits come first, followed by the less significant bits in order of significance. Picking out text that is layered on top of a background image is relatively easy with wavelets: just pick out the really large coefficients in the wavelet expansion, since those most likely correspond to parts of the image where large jump discontinuities occur. This can all be done automatically, of course.
Clustering IS STILL brute-force computing. All clustering does is make a more powerful computer; but it can't do any computations that a serial computer can't do. The only advantage clustering offers is more computing power, which is the same advantage that you get by using a faster serial computer.
This stuff sounds a lot like the science fiction of the forties and fifties..."By 1980, people will travel to their offices in automated helicopters! People will fly to the moon and mars on a regular basis!" Many science fiction novels of the time portrayed extravagant space-travel, and yet had humans doing the navigational computations on board these spacecraft, completely missing the coming dominance of electronic computers in computations. The popular imagination of future technology was simply an extrapolation of the current technology - "Fast airplanes, bigger rockets, atomic power!" - when in reality the technology of the future has turned out to be things that people could not have imagined at the time (the world-wide-web, biotechnology, fax machines, massive computational power...) In the same vein, the breakthrough technology of the next 100 years will most likely consist of things that haven't even occured to us yet -- not simply faster and faster computers, smaller and smaller robots, better and better bioengineering, but rather something else entirely.
In grad school I took a unix systems administration class from Evi Nemeth (one of the authors of the book). This was 5 years ago, so we used the Red version of the book. While the book was indeed very useful, I found and continue to find that I have to look at man pages or other info on the web if I really want to know the details of how to do something with a system. Also, I think she didn't like me, because unlike the undergrads in the class, I was willing to challenge her on things (like, "Why are you always late to class, professor?")
The point of the neural net these people have set up is not to have a neural net that recognizes words well; nor is it to have a neural net to recognize words in noisy environments from different speakers. Rather, the point of the research is to study how biological neurons can process sensory information over time to produce a sensation of a "moment." Recognizing the word "one" is an example of such temporal sensory processing, and a relatively easy one to test, so that's why they chose that task for their research project.
Mac OS X is isn't totally unrelated to MkLinux. MkLinux is based on the Mach microkernel, and Apple leveraged their development on the MkLinux project to improve the Mach kernel they are using in Mac OS X.
That is an interesting observation that you made about education level correlating with the use of "HA HA HA." As it happens, I have a Ph.D., so I guess I must be an outlier in your study.
Collage carrier! HA HA HA !! That is one of the funniest mis-spellings I've ever seen on slashdot. You may be a college student, but you sure aren't very well educated. (Collage carrier! That sounds like a special rack designed for carrying miniature flower bouquets).