As a general, wide-sweeping software design idea, I think scriptable programs are a Good Thing. This pretty well necessitates "executable documents" assuming that you want to carry macros back and forth. The problem is with invisible, automatically-executing macros (or invisible, automatically-executing code in general). Adding more functionality to a program (a la Emacs) really helps in customization, and is a good design choice. The problems of word macros are the same design flaws that lead to insecure operating systems and programming languages.
Okay, as far as I can tell, this guy posted some theories about security loopholes peculiar to opensource systems. He was flamed, and was put off by, in his eyes, the arrogance and short-sightedness of free software advocates. In his opinion, his security ideas were not given fair examination.
So I read his article. I found the prose a little clumsy, but as best I can tell his argument is as follows:
1) Having the source code allows you to alter software in very subtle ways, and recompile it 2) Installing this new software could be so subtle as to remain undetected by any administrators
At least I think that's his argument. My challenges to this theory would be: How is J. Random User able to reinstall altered software? Why would open source code pose such a security loophole, since machine code (being a programming language) can be reprogrammed by savvy coders just as easily (or easier!) than source code could?
If the author is reading this, I would recommend him to www.counterpane.com to Bruce Schneier's writings. Schneier, as far as I know, does not give a rat's ass about RMS's vision of a world free of proprietary software. But Schneier is fanatical about security systems and cryptographic protocols undergoing peer review, and the need to avoid "security by obscurity." Machine code can be reprogrammed. Machine code is a programming language. Just because code isn't easy to change doesn't mean it *can't* be changed.
I may well have misunderstood his arguments; if so, I apologize. It sounded like this dude is unused to getting flamed. (Who doesn't get mad at being flamed?) But that shit happens all the time in a public forum, and it happens to absolutely everyone who posts.
It's good that you've taken MS's FUD to heart
on
iMac Linux
·
· Score: 1
This sort of "oh it's outdated, oh it's like cutting your lawn with a scythe, oh it's like running an 8088" is the kind of thing that shows up from pro-Windows users.
And did you actually read my post? My point was that when the iMac becomes unusably obselete as a workstation (even if you think that's the case right now), it will still be well-suited to be an X terminal. It's better suited to be an X terminal than most other computers. Such as, say, a compaq presario or other random PC, because it has a good monitor, a combined monitor/cpu, and because it's neat looking.
I think I've come to not understand what people mean by "flames." I almost never see posts that have no redeeming value whatsoever. Most posts that get categorized as flames are contreversial or ill-worded, not worthless. A lot of what Linus said in the article (microkernels are a joke, Emacs sucks, gcc is the only really good GNU tool) would surely get him a low score on/.
I ordered mine in early february, and even though I got what apparently is the least popular color (tangerine!), it still took like a month from the time they cashed my check to the time they shipped it. That was direct from apple, though. But I hear nowadays that everyone is out of all the new ones. Good luck.
Two words: X Terminal. (One word?)
on
iMac Linux
·
· Score: 1
iMacs are perfect as X terminals (well, not perfect, but pretty good). They're compact, they have very good monitors, they have more than enough resources to run an X server cleanly. Granted, you can do more than just run X on the things. But when Moore's law knocks down *any* machine you buy, you'll want something that still has some use while obselete.
(Sorry about continuing the off-topic stuff, but the shooting pains in my hands made the previous post stand out.)
Do you know of any good resources for rolling one's own keymap? The keyboard / terminal HOWTO was less than totally thorough, IMHO. I was especially intrigued by the (patented) half-QWERTY keyboard mentioned in the article. The design sounds easy-to-learn, but how does one make the space bar sticky?
I'd be happy with an RTFM, as long as it pointed to a good FM.
This, of course, brings up another issue: Is there a minimum integer that/. uses? At what point does my incredibly low threshold level become an incredibly high threshold level? Or is some limitless integer type, or is there some threshold that translates to NaN?
I wish I could stop obsessing about these things. Too bad it's my job.:)
I may be on crack, but the implicit "sending others the money you make on your modified version of the software" sounds a lot like a pyramid scheme.
I start the scheme, I write some software program and license it under the NCL. I charge $2 to sell it to others. Alice buys the software, makes a change to it, sells it for $2. She, by the license, sends back a portion of her money to me. She sells Bob a copy. Bob makes a change, sells his copy. Lather, rinse, repeat.
Okay. Here, to the best of my understanding, is a history of GNU.
A long time ago, Richard Stallman was a student at MIT. At that time, all software was free and open, and RMS really dug that. Then he ran across some proprietary software, and it pissed him off that it was proprietary.
It really pissed him off.
It REALLY REALLY pissed him off.
So much so that he dedicated his life to replacing all proprietary software with free alternative versions.
He decided that Unix was a pretty popular thing, so he sat down to make a free version of Unix. He called it GNU, recursively meaning "GNU's Not Unix," because it wasn't Unix, it was a free clone of Unix. (Unix was owned by AT&T. Currently, as a bit of trivia, the name "Unix" is owned by SCO.)
So GNU was supposed to be a full-fledged free Unix clone. That was the idea. RMS and company started making a lot of software, leaving the kernel until last for some reason. All the other GNU OS components were pretty cool, such as EMACS and GCC and such. So people would use them in proprietary versions of Unix. But still RMS's dream was to make a free system, GNU.
Then Linus Torvalds, in 91, decides to fuck around with making a kernel based on Andy Tanenbaum's MINIX. Because, hey, don't we all want to make our own kernel.
So he makes his own kernel which he calls "Linux." People convince him to make it as compatible as possible with the GNU utilities, and to license it under the GNU license.
Keep in mind that at this point, GNU (RMS's free Unix clone) is almost done except for a kernel. Linux is released, and bam, fills in the missing link.
So from RMS's point of view, it's as if this decades-long project -- his life's work, really -- was taken away from him at the last second. This work of his, GNU, this free version of Unix he had planned out, was being called Linux by everyone because that was the kernel it was based on.
This, it should be noted, is hilarious. The whole idea of free software is that it isn't owned by anyone. In such a system, it is necessarily difficult to give credit where credit is due. So RMS is mad at his own creation -- unownable software.
Linus Torvalds, as far as I can tell, could not care less about who gets the credit, or what the damn thing is named, or whether all proprietary software is killed or not. He mostly seems to care about seeing that his family is well-cared for and that he has a stimulating job.:)
I thought it was a misuse of the term to say "bible" for the holy text of any religion, but Webster's dictionary backs you up.
From dictionary.com:
bible \Bi"ble\ (b[imac]"b'l), n. [F. bible, L. biblia, pl., fr. Gr. bibli`a, pl. of bibli`on, dim. of bi`blos, by`blos, book, prop. Egyptian papyrus.] 1. A book. [Obs.] --Chaucer.
2. The Book by way of eminence, -- that is, the book which is made up of the writings accepted by Christians as of divine origin and authority, whether such writings be in the original language, or translated; the Scriptures of the Old and New Testaments; -- sometimes in a restricted sense, the Old Testament; as, King James's Bible; Douay Bible; Luther's Bible. Also, the book which is made up of writings similarly accepted by the Jews; as, a rabbinical Bible.
3. A book containing the sacred writings belonging to any religion; as, the Koran is often called the Mohammedan Bible. Source: Webster's Revised Unabridged Dictionary
Or as Orwell wrote, "Freedom is Slavery." To be free is to be a slave, there is no difference between the two. Larry Flynt is free to write what he wants but is a prisoner to his wheelchair as a result. Copylefted software remains free by forcing new contributions to also be GPLed. We send kids to fight in wars to protect "freedom."
As for whether deconstructions and the big picture are part of the same thing, I'm not sure I agree. Deconstructionism, like any metacognitive tool, has a tendancy to spin off to nowhere when people use it to analyze itself. (Like doublethink, or when the people in "White Noise" go to take pictures of the most-photographed barn in the world just because it is the most-photographed barn in the world.) The end result is that nothing is trusted, that there is no "fact." Which is okay philisophically, but in my experience lends itself to shallow social analysis. Deconstruction just seems overused in many academic discourses these days.
It's spelled "Foucault." (I misspelled it in an earlier post as well.)
I think this article is actually being flamed a lot less than it might in other circles. I've heard those from a literary / social science background really roast people who use terms like "commodification" and "social construction" to cover up a meaningless argument.
I'm a computer scientist, but went through a liberal arts college and thus got a lot of the social sciences mixed in with the algorithms.
Deconstruction (structuralism) was largely created by Michel Foucalt in the 60s as a reaction to people spewing out facts without critically analyzing them. One his best books, "Madness and Civilization," deconstructs the rather wishy-washy idea of "insanity," and how the idea that people are "insane" while others are "sane" has its roots in social factors as well as science.
The deconstruction works really well for social concepts which are widely believed but are really crap. Racist beliefs, sexist beliefs, these are all begging to be deconstructed.
But the problem, as Noam Chomsky said, is that social science has a lot of bredth but not much depth. So once someone comes up with a good idea, everyone hops on the bandwagon and uses that idea to analyze *everything.* This happened with Freudianism. After Freud's work came out, literary analysts would imagine that everything in, say, the "Iliad" or Kafka's "Metamorphosis" was all about sex. (Once you get a hammer, everything is a nail.) Deconstruction is useful in a lot of discussions, but social scientists are using it everywhere (i.e., to analyze computer hacking). It's like the social scientist in Don Delillo's "White Noise" who devotes his career to analyzing cereal boxes.
But if you're really interested, pick up "Madness and Civilization," or one like "Contested Lives" (deconstructing the abortion debate). [Vine Deloria's "Custer Died for your Sins" is good too, and reads like a deconstruction although Deloria probably didn't intend it to be so.]
I liked his article until he started using terms like "social construction" and "deconstruction." I like Foucalt as much as the next guy, but why does EVERY SINGLE ACADEMIC SOCIAL COMMENTARY have to be a poststructuralist deconstruction? It's a useful discussion tool, but it's overused and cliche these days. At least it was in the sociology and anthropology courses I took in college.
I liked him at first, but now I say ditch him. He's dependant on buzz words.
I've heard RMS's take on MP3s, and the fear he expressed was that it would be outlawed by the government it ways that strong encryption exporting has been outlawed. (It's not that no one exports strong encryption, it just means that the makers of software with strong encryption in it have to obey the stupid export laws.)
This is, of course, a far cry from the nonsense that ZDNet was spewing ("Microsoft will take over MP3's!"). But with regards to your post, the use of MP3s, or any software the government dislikes, is not a certain freedom.
I've got the second edition here on my desk, in paperback form. I purchased it while still in college (I graduated May'98, probably purchased the book about a year prior). The only copyright date on the inside is listed as 1996.
I think there are some different wavelengths going on here. While many writers have said that, for them, the medium in which writing occurs affects the content of what is written, this does not exclude certain media from possibly being used to produce "art" or "professional writing." I can think of many textbooks which have been written on Unices, several of which would qualify as good professional writing (the rest of which would qualify as crap, grumble grumble). While some writers prefer to write in pencil rather than pen, on typewriter rather than computer, or in sand rather than Windows98, that does not mean that quality work *cannot* be produced in these media.
Perhaps your question is more along the lines of, "do any non-technical professional writers prefer to write their stuff on computers running Linux rather than other media?"
The vehemence with which the right wing goes after Clinton has always baffled me. He's probably the most conservative Democrat since the "solid south" days.
Slashdot Mirror
As a general, wide-sweeping software design idea, I think scriptable programs are a Good Thing. This pretty well necessitates "executable documents" assuming that you want to carry macros back and forth. The problem is with invisible, automatically-executing macros (or invisible, automatically-executing code in general). Adding more functionality to a program (a la Emacs) really helps in customization, and is a good design choice. The problems of word macros are the same design flaws that lead to insecure operating systems and programming languages.
Okay, as far as I can tell, this guy posted some theories about security loopholes peculiar to opensource systems. He was flamed, and was put off by, in his eyes, the arrogance and short-sightedness of free software advocates. In his opinion, his security ideas were not given fair examination.
So I read his article. I found the prose a little clumsy, but as best I can tell his argument is as follows:
1) Having the source code allows you to alter software in very subtle ways, and recompile it
2) Installing this new software could be so subtle as to remain undetected by any administrators
At least I think that's his argument. My challenges to this theory would be: How is J. Random User able to reinstall altered software? Why would open source code pose such a security loophole, since machine code (being a programming language) can be reprogrammed by savvy coders just as easily (or easier!) than source code could?
If the author is reading this, I would recommend him to www.counterpane.com to Bruce Schneier's writings. Schneier, as far as I know, does not give a rat's ass about RMS's vision of a world free of proprietary software. But Schneier is fanatical about security systems and cryptographic protocols undergoing peer review, and the need to avoid "security by obscurity." Machine code can be reprogrammed. Machine code is a programming language. Just because code isn't easy to change doesn't mean it *can't* be changed.
I may well have misunderstood his arguments; if so, I apologize. It sounded like this dude is unused to getting flamed. (Who doesn't get mad at being flamed?) But that shit happens all the time in a public forum, and it happens to absolutely everyone who posts.
This sort of "oh it's outdated, oh it's like cutting your lawn with a scythe, oh it's like running an 8088" is the kind of thing that shows up from pro-Windows users.
And did you actually read my post? My point was that when the iMac becomes unusably obselete as a workstation (even if you think that's the case right now), it will still be well-suited to be an X terminal. It's better suited to be an X terminal than most other computers. Such as, say, a compaq presario or other random PC, because it has a good monitor, a combined monitor/cpu, and because it's neat looking.
I think I've come to not understand what people mean by "flames." I almost never see posts that have no redeeming value whatsoever. Most posts that get categorized as flames are contreversial or ill-worded, not worthless. A lot of what Linus said in the article (microkernels are a joke, Emacs sucks, gcc is the only really good GNU tool) would surely get him a low score on /.
I ordered mine in early february, and even though I got what apparently is the least popular color (tangerine!), it still took like a month from the time they cashed my check to the time they shipped it. That was direct from apple, though. But I hear nowadays that everyone is out of all the new ones. Good luck.
iMacs are perfect as X terminals (well, not perfect, but pretty good). They're compact, they have very good monitors, they have more than enough resources to run an X server cleanly. Granted, you can do more than just run X on the things. But when Moore's law knocks down *any* machine you buy, you'll want something that still has some use while obselete.
(Sorry about continuing the off-topic stuff, but the shooting pains in my hands made the previous post stand out.)
Do you know of any good resources for rolling one's own keymap? The keyboard / terminal HOWTO was less than totally thorough, IMHO. I was especially intrigued by the (patented) half-QWERTY keyboard mentioned in the article. The design sounds easy-to-learn, but how does one make the space bar sticky?
I'd be happy with an RTFM, as long as it pointed to a good FM.
This, of course, brings up another issue: Is there a minimum integer that /. uses? At what point does my incredibly low threshold level become an incredibly high threshold level? Or is some limitless integer type, or is there some threshold that translates to NaN?
:)
I wish I could stop obsessing about these things. Too bad it's my job.
Nor does it cover the timeout-to-GPL thing in the NCL.
I may be on crack, but the implicit "sending others the money you make on your modified version of the software" sounds a lot like a pyramid scheme.
I start the scheme, I write some software program and license it under the NCL. I charge $2 to sell it to others. Alice buys the software, makes a change to it, sells it for $2. She, by the license, sends back a portion of her money to me. She sells Bob a copy. Bob makes a change, sells his copy. Lather, rinse, repeat.
Time to write that NCLed "Hello World" program.
Okay. Here, to the best of my understanding, is a history of GNU.
:)
A long time ago, Richard Stallman was a student at MIT. At that time, all software was free and open, and RMS really dug that. Then he ran across some proprietary software, and it pissed him off that it was proprietary.
It really pissed him off.
It REALLY REALLY pissed him off.
So much so that he dedicated his life to replacing all proprietary software with free alternative versions.
He decided that Unix was a pretty popular thing, so he sat down to make a free version of Unix. He called it GNU, recursively meaning "GNU's Not Unix," because it wasn't Unix, it was a free clone of Unix. (Unix was owned by AT&T. Currently, as a bit of trivia, the name "Unix" is owned by SCO.)
So GNU was supposed to be a full-fledged free Unix clone. That was the idea. RMS and company started making a lot of software, leaving the kernel until last for some reason. All the other GNU OS components were pretty cool, such as EMACS and GCC and such. So people would use them in proprietary versions of Unix. But still RMS's dream was to make a free system, GNU.
Then Linus Torvalds, in 91, decides to fuck around with making a kernel based on Andy Tanenbaum's MINIX. Because, hey, don't we all want to make our own kernel.
So he makes his own kernel which he calls "Linux." People convince him to make it as compatible as possible with the GNU utilities, and to license it under the GNU license.
Keep in mind that at this point, GNU (RMS's free Unix clone) is almost done except for a kernel. Linux is released, and bam, fills in the missing link.
So from RMS's point of view, it's as if this decades-long project -- his life's work, really -- was taken away from him at the last second. This work of his, GNU, this free version of Unix he had planned out, was being called Linux by everyone because that was the kernel it was based on.
This, it should be noted, is hilarious. The whole idea of free software is that it isn't owned by anyone. In such a system, it is necessarily difficult to give credit where credit is due. So RMS is mad at his own creation -- unownable software.
Linus Torvalds, as far as I can tell, could not care less about who gets the credit, or what the damn thing is named, or whether all proprietary software is killed or not. He mostly seems to care about seeing that his family is well-cared for and that he has a stimulating job.
www.fsf.org for more info.
I thought it was a misuse of the term to say "bible" for the holy text of any religion, but Webster's dictionary backs you up.
From dictionary.com:
bible \Bi"ble\ (b[imac]"b'l), n. [F. bible, L. biblia, pl., fr. Gr. bibli`a, pl. of bibli`on, dim. of bi`blos, by`blos, book, prop.
Egyptian papyrus.] 1. A book. [Obs.] --Chaucer.
2. The Book by way of eminence, -- that is, the book which is made up of the writings accepted by Christians as of divine
origin and authority, whether such writings be in the original language, or translated; the Scriptures of the Old and New
Testaments; -- sometimes in a restricted sense, the Old Testament; as, King James's Bible; Douay Bible; Luther's Bible.
Also, the book which is made up of writings similarly accepted by the Jews; as, a rabbinical Bible.
3. A book containing the sacred writings belonging to any religion; as, the Koran is often called the Mohammedan Bible.
Source: Webster's Revised Unabridged Dictionary
Or as Orwell wrote, "Freedom is Slavery." To be free is to be a slave, there is no difference between the two. Larry Flynt is free to write what he wants but is a prisoner to his wheelchair as a result. Copylefted software remains free by forcing new contributions to also be GPLed. We send kids to fight in wars to protect "freedom."
As for whether deconstructions and the big picture are part of the same thing, I'm not sure I agree. Deconstructionism, like any metacognitive tool, has a tendancy to spin off to nowhere when people use it to analyze itself. (Like doublethink, or when the people in "White Noise" go to take pictures of the most-photographed barn in the world just because it is the most-photographed barn in the world.) The end result is that nothing is trusted, that there is no "fact." Which is okay philisophically, but in my experience lends itself to shallow social analysis. Deconstruction just seems overused in many academic discourses these days.
It's spelled "Foucault." (I misspelled it in an earlier post as well.)
I think this article is actually being flamed a lot less than it might in other circles. I've heard those from a literary / social science background really roast people who use terms like "commodification" and "social construction" to cover up a meaningless argument.
I'm a computer scientist, but went through a liberal arts college and thus got a lot of the social sciences mixed in with the algorithms.
Deconstruction (structuralism) was largely created by Michel Foucalt in the 60s as a reaction to people spewing out facts without critically analyzing them. One his best books, "Madness and Civilization," deconstructs the rather wishy-washy idea of "insanity," and how the idea that people are "insane" while others are "sane" has its roots in social factors as well as science.
The deconstruction works really well for social concepts which are widely believed but are really crap. Racist beliefs, sexist beliefs, these are all begging to be deconstructed.
But the problem, as Noam Chomsky said, is that social science has a lot of bredth but not much depth. So once someone comes up with a good idea, everyone hops on the bandwagon and uses that idea to analyze *everything.* This happened with Freudianism. After Freud's work came out, literary analysts would imagine that everything in, say, the "Iliad" or Kafka's "Metamorphosis" was all about sex. (Once you get a hammer, everything is a nail.) Deconstruction is useful in a lot of discussions, but social scientists are using it everywhere (i.e., to analyze computer hacking). It's like the social scientist in Don Delillo's "White Noise" who devotes his career to analyzing cereal boxes.
But if you're really interested, pick up "Madness and Civilization," or one like "Contested Lives" (deconstructing the abortion debate). [Vine Deloria's "Custer Died for your Sins" is good too, and reads like a deconstruction although Deloria probably didn't intend it to be so.]
I liked his article until he started using terms like "social construction" and "deconstruction." I like Foucalt as much as the next guy, but why does EVERY SINGLE ACADEMIC SOCIAL COMMENTARY have to be a poststructuralist deconstruction? It's a useful discussion tool, but it's overused and cliche these days. At least it was in the sociology and anthropology courses I took in college.
I liked him at first, but now I say ditch him. He's dependant on buzz words.
I've heard RMS's take on MP3s, and the fear he expressed was that it would be outlawed by the government it ways that strong encryption exporting has been outlawed. (It's not that no one exports strong encryption, it just means that the makers of software with strong encryption in it have to obey the stupid export laws.)
This is, of course, a far cry from the nonsense that ZDNet was spewing ("Microsoft will take over MP3's!"). But with regards to your post, the use of MP3s, or any software the government dislikes, is not a certain freedom.
I've got the second edition here on my desk, in paperback form. I purchased it while still in college (I graduated May'98, probably purchased the book about a year prior). The only copyright date on the inside is listed as 1996.
Why choose now to review it? Maybe I'm just disappointed that there isn't a new edition out. Great book.
Although the food sucks
Rob! Fix the plain text stuff!
The "less than" character gets eaten, even in "Plain Old Text" mode. Here we go in C, then:
#include
int main (void) {
printf ("English is too confusing. ");
printf ("Try writing a book in C.\n");
return (0);
}
It's C++, and incorrect C++ at that. Also, the string is weirdly formatted and grammatically incorrect. Try this:
#include "iostream.h"
int main (void) {
cout }
I think there are some different wavelengths going on here. While many writers have said that, for them, the medium in which writing occurs affects the content of what is written, this does not exclude certain media from possibly being used to produce "art" or "professional writing." I can think of many textbooks which have been written on Unices, several of which would qualify as good professional writing (the rest of which would qualify as crap, grumble grumble). While some writers prefer to write in pencil rather than pen, on typewriter rather than computer, or in sand rather than Windows98, that does not mean that quality work *cannot* be produced in these media.
Perhaps your question is more along the lines of, "do any non-technical professional writers prefer to write their stuff on computers running Linux rather than other media?"
The vehemence with which the right wing goes after Clinton has always baffled me. He's probably the most conservative Democrat since the "solid south" days.