Ho, ho, don't get all up in arms you guys. Very likely the people here claiming the mormon religion is "crazy" would, when pressed, also claim any of a great number of other religions to be "crazy". To people for whom critical evaluation of fallible hypotheses is more the "obvious" thing to do then to accept a certain amount of dogma as completely impervious to reason, any religion that embodies arbitrary certainties is suspect.
Now try to imagine your self in such a sceptical frame of reference, and then compare the Latter Day Saints to the Catholics. And now imagine your trying to imagine how anyone could believe in either religion. Then you may find that the more recent a religion is, and hence the more insight we have into its origins, the harder it gets to understand that its associated group of believers don't "see it's a fraud".
None of this has anything to do with how far your religion's (perceived) social and cultural norms -- as opposed to theological doctrine -- deviate from those present in the general community. I don't think it's fair to say that anyone who doesn't like the mormon religion is just the victim of some FUD about polygamy.
Over here, we had the "silverware computer" running a webserver for a while. It was a standard AMD box assembled out of mostly dumpster-dived components, except instead of giving it a proper case, they stuck it in a wooden kitchen drawer. It served webpages just fine as long as nobody closed the drawer all the way, which would cause the AMD to overheat.
You won't, for the same reason you won't invade China. It's bad for the US economy.
Iraq was never a major customer for US exports, hell, they were mostly a major supplier of US imports (oil). By essentially colonizing them, you suddenly get a market for US exports (Haliburton), while at the same time the oil will keep flowing to the US just fine. Destroying some portion of the production infrastructure is not a problem, since Iraq's main asset is the oil itself rather than the pumps and pipes.
The EU, like China, have a different kind of economic relationship to the US. There is intensive trade, both ways, in highly value-added products (computers, cars, whatnot). A war would put a sudden stop to that for at least its duration, but probably much longer, since the complicated and fragile infrastructure needed for the sort of economy on hand is very easily destroyed by warfare.
IANAL, but for all I know the European Court of Justice can base its decisions solely on some Charter of Rights. It's not very likely Microsoft will get anything out of that, since the Charter is (and rightfully so) mostly biased towards individual citizens and doesn't really extend much in the way of rights to corporations. Of course, there may be some "interesting" precedent somewhere, but it still seems like a meager basis for challenging an essentially administrative (rather than judicial) anti-trust decision.
The pricing *does* have to do with the weak dollar. Not in the sense of "the weak dollar made the US economy go bad and now they have to dump everything at low prices", but in the sense of "I have 1000 Euro to buy a computer, two months ago that was $900, now it's suddenly $1220". Of course the same holds for the EU retailers, but they tend to respond very slowly to currency changes. And then there's also a discrepancy in VAT: in much of EU, VAT on non-essential items approaches 20%, in the US it can be as low 0%, and very uncommonly more than 8%.
As for import and export. I have travelled between California and Europe back and forth about a dozen times over the last few years. Some of the time I "lived" in California, some of the time I "lived" in Europe, and some of the time it was unclear. I have never been asked to pay tax on a computer. The thing is that I never brought one in a sealed cardboard box screaming "Brand New Macintosh!". It was always sitting in my backpack, dvorak keys scrawled on with a Sharpie, decorated with a sticker advertising some political cause, in short, I was clearly just bringing it along because I *use* the thing. I took it out of the bag for the X-rays in the US, left it in for the X-rays in the EU, just like everyone else. My father travels for business almost monthly and for all I know he has the same experience as I do, but then again all the "BusinessElite" tags may help him a bit.
I do not know what this is like in some other parts of Europe, but in The Netherlands people have always been perfectly happy buying English-language software if nothing localized was available.
Which is a good thing, because with everything you can say for and against the Linux kernel, at least it will run on a wide variety of processors, some of which are one HELL of a lot better designed than the x86. The MIPS is a textbook example (literally) of a RISC processor, which lends itself perfectly to heavy pipelining and the like, thus making a 500 MHz MIPS a lot more impressive than it sounds. Also, the particular MIPS flavor the Chinese are working on is 64-bits, which gives them another means to approach the performance of the western chips in at least some applications without quite making the same clock speed.
I have done considerable mathematical editing in WordPerfect 5.1 and 6.0 for DOS. It was more bearable than the Microsoft Equation Editor, which I later *tried* to use, since there was some sort of TeX-like notation you could use. It got really, really messy when I tried to typeset a term paper on head-driven phrase structure grammar (an interesting branch of theoretical linguistics originated at Stanford).
These days I use (La)TeX, whether there are equations in my document or not. Like emacs, or even WP 5.1, or for that matter just about any computer program that actually *increases* my productivity, it took a while to learn, and I still quite frequently refer to the Not So Short Guide and the Gentle Introduction, but it really pays off, because it's a well-designed system that wasn't shipped out in a rush by a software company that thinks ease of learning equals ease of use and employs programmers who have never opened a CS textbook.
Ho, ho, slow down... Music is a very pluriform sort of thing, that comes in many different types. Some people maybe into musical traditions that do not, traditionally, view standard music notation as an absolute requirement, such as pop, folk, or rock. Often, the use of improvisation and a more fluid means of adapting each other's work lend great charm to these genres, and I would be the last to condemn a good bit of folk, rock, blues or what-have-you.
At the same time, plenty of people immensely enjoy types of music where notation is very important. I, for one, am a great admirer and mediocre practitioner of western classical music, a particularly rich musical tradition that dates back to the middle ages. Within that particular tradition, musical notation still plays a very important role. Some of the reasons are that: - it removes the need to memorize lengthy and complicated pieces before one can play them at all (although good musicians admittedly have a tendency to memorize anyhow) - it can serve as an abstract medium of expression for a composer that maybe doesn't even play the instrument(s) s/he's writing for - it can help further the ideal of playing music "as the composer intended", which has traditionally been valued highly, and only more so recently, by allowing a separation of the music from any particular performance - it provides a medium for the theoretical study of musical patterns - it provides a cheap and accessible way to catalogue and refer to the immense body of music composed over the centuries - symphonic and choral pieces require immense amounts of coordination between large groups of musicians, which would be pretty much undoable without some form of notation &c.
The fact that he's going to be charged in Australia anyway, which may very well be true, is not that relevant. The premise behind your excuse is that his situation wouldn't be any different on either side of the pacific. I would beg to differ.
In the US, prison sentences are often longer than in other places, and prison conditions are atrocious to the point where violence, rape and murder between inmates are barely considered abnormal anymore. Trials are very heavily influenced by the amount of money you can spend on a lawyer. I have no particular knowledge of Australia, but I would imagine the guy's situation could be considerably different if he were tried at home.
What exactly do you mean by "updated for modern functional languages"? The main functional languages these days are Common Lisp, Scheme, OCaml, SML, and Haskell.
The original LISP machines ran Zetalisp, which had a lot of Common Lisp in it already, but still had quirks like dynamic scope. Most Lisp Machines derived from the MIT lineage (Symbolics, LMI and TI) supported Common Lisp as soon as it came around; Xerox had a competing system that ran Interlisp.
Now, if by "modern" functional language you mean ANSI Common Lisp or Scheme, I agree: it would be really awesome to have a modern Lisp Machine around, hopefully a little more RISC-like, but with the same basic philosophy. If you mean some ML dialect, though, the use of special-purpose architecture seems limited to me. The main win of the Lisp machines was that they could deal with typed values; in ML you don't need those, since types are dealt with statically.
Lastly, let me just shamelessly promote my Lisp Machine webpage:
Given proper tail recursion, goto's are completely equivalent to function calls.
Look at the following example, written in an imaginary C-like language
start:
twiddle();
if (...) {goto f; } else {goto g;}
f:
tweedle();
g:
twaddle();
goto f
Now with function calls (I'm going to use Scheme syntax here, but don't be scared):
(define (start)
(twiddle)
(if... (f) (g)))
(define (f)
(tweedle)
(g))
(define (g)
(twaddle)
(f))
You may *think* that every function call is (a) slow and (b) pushes stack, but that's not necessarily the case. The standard paper (good read!) is ftp://publications.ai.mit.edu/ai-publications/pdf/ AIM-443.pdf .
It's pronounced "oh camel", i.e. *not* as a fictitious Irish surname but as an analogue to ACaml, BCaml, CCaml...
The big difference between OCaml and Lisp/Scheme is that even though both are *strongly* typed, that is, you can never, say, reference memory address pi becaus a float is simply not a pointer, they work it in different ways. Lisp/Scheme is dynamically typed, like Python, and stores the types with the values, which are checked at runtime. OCaml is statically typed, which means it does not need to store types in tags, but analyzes your program at compile time to ensure "type safety".
On the one hand, this gives a performance win. On the other hand, it makes programming slightly more painful at times. The expression "if x=3 then 5 else 2.0" is not acceptable to the OCaml compiler, because it cannot determine its type; the expression (if (eq? x 3) 5 2.0) is perfectly acceptable to a Scheme compiler. The advantage of this, it is claimed, is that most programs without well-defined types are buggy anyhow, so it helps you write good code.
Note that the same performance win can be had from not having the language be very precise about types at all, and randomly allowing dereferencing of integers and multiplication of pointers. The archetype for that, of course, is C, which is considered *weakly* typed (a.k.a. "wrongly typed"). This has the disadvantage that programs can behave unpredictably and cause system crashes very easily. If an OCaml or Scheme program ever segfaults, let alone causes a bus error, that's a bug in the compiler. If a C program does that, well you all know what that means...
Just to confuse you a little more, there's yet another distinction to be made, which is between explicitly and implicitly typed. A language such as C or Java is explicitly typed, meaning that you have to declare variables and functions as being of a certain type, although the explicit types in C are mostly sugar and can always be overridden. Python and Scheme are implicitly typed, that is, you don't mention any types, which is not a big deal, since there really is only one type. All right, backtrack. I just said they had type tags, and now I say there's only one type? The answer to that is one of terminology. From a theoretical perspective, if any slot can contain any value, there's only one type (like Java's Object).
Now OCaml is statically, strongly *and* implicitly typed. That means that it will check at compile time that there can be no type errors, and you don't even have to mention the types. Say you write a funtion in OCaml like this: let f x = 2 * x . Then it will derive for you that it has type Int -> Int. This is accomplished through a piece of higher math called Hindley-Milner type inferencing.
If you are interested in this sort of stuff, I strongly recommend Lambda the Ultimate (lambda.weblogs.com), which is always full of programming language reading material, both on a very abstruse theoretical and on a very practical level.
That is definitely not the case in the US. If anyone wants to know my grades, they have to ask me, and then I can request a transcript to be sent to them by the Institute. Exceptions are made for Institute officials, but I had to explicitly sign a form authorizing them to release them to scholarship organizations. From the time I spent at UC Berkeley, I seem to remember the same is true for public universities.
Slashdot Mirror
Ho, ho, don't get all up in arms you guys. Very likely the people here claiming the mormon religion is "crazy" would, when pressed, also claim any of a great number of other religions to be "crazy". To people for whom critical evaluation of fallible hypotheses is more the "obvious" thing to do then to accept a certain amount of dogma as completely impervious to reason, any religion that embodies arbitrary certainties is suspect.
Now try to imagine your self in such a sceptical frame of reference, and then compare the Latter Day Saints to the Catholics. And now imagine your trying to imagine how anyone could believe in either religion. Then you may find that the more recent a religion is, and hence the more insight we have into its origins, the harder it gets to understand that its associated group of believers don't "see it's a fraud".
None of this has anything to do with how far your religion's (perceived) social and cultural norms -- as opposed to theological doctrine -- deviate from those present in the general community. I don't think it's fair to say that anyone who doesn't like the mormon religion is just the victim of some FUD about polygamy.
Over here, we had the "silverware computer" running a webserver for a while. It was a standard AMD box assembled out of mostly dumpster-dived components, except instead of giving it a proper case, they stuck it in a wooden kitchen drawer. It served webpages just fine as long as nobody closed the drawer all the way, which would cause the AMD to overheat.
You won't, for the same reason you won't invade China. It's bad for the US economy.
Iraq was never a major customer for US exports, hell, they were mostly a major supplier of US imports (oil). By essentially colonizing them, you suddenly get a market for US exports (Haliburton), while at the same time the oil will keep flowing to the US just fine. Destroying some portion of the production infrastructure is not a problem, since Iraq's main asset is the oil itself rather than the pumps and pipes.
The EU, like China, have a different kind of economic relationship to the US. There is intensive trade, both ways, in highly value-added products (computers, cars, whatnot). A war would put a sudden stop to that for at least its duration, but probably much longer, since the complicated and fragile infrastructure needed for the sort of economy on hand is very easily destroyed by warfare.
IANAL, but for all I know the European Court of Justice can base its decisions solely on some Charter of Rights. It's not very likely Microsoft will get anything out of that, since the Charter is (and rightfully so) mostly biased towards individual citizens and doesn't really extend much in the way of rights to corporations. Of course, there may be some "interesting" precedent somewhere, but it still seems like a meager basis for challenging an essentially administrative (rather than judicial) anti-trust decision.
The pricing *does* have to do with the weak dollar. Not in the sense of "the weak dollar made the US economy go bad and now they have to dump everything at low prices", but in the sense of "I have 1000 Euro to buy a computer, two months ago that was $900, now it's suddenly $1220". Of course the same holds for the EU retailers, but they tend to respond very slowly to currency changes. And then there's also a discrepancy in VAT: in much of EU, VAT on non-essential items approaches 20%, in the US it can be as low 0%, and very uncommonly more than 8%.
As for import and export. I have travelled between California and Europe back and forth about a dozen times over the last few years. Some of the time I "lived" in California, some of the time I "lived" in Europe, and some of the time it was unclear. I have never been asked to pay tax on a computer. The thing is that I never brought one in a sealed cardboard box screaming "Brand New Macintosh!". It was always sitting in my backpack, dvorak keys scrawled on with a Sharpie, decorated with a sticker advertising some political cause, in short, I was clearly just bringing it along because I *use* the thing. I took it out of the bag for the X-rays in the US, left it in for the X-rays in the EU, just like everyone else. My father travels for business almost monthly and for all I know he has the same experience as I do, but then again all the "BusinessElite" tags may help him a bit.
Apple is no deader than usual.
"Hey, where would Microsoft get all their R&D from if not for Apple?" From late-60s era Xerox PARC systems, like everyone else.
I do not know what this is like in some other parts of Europe, but in The Netherlands people have always been perfectly happy buying English-language software if nothing localized was available.
Which is a good thing, because with everything you can say for and against the Linux kernel, at least it will run on a wide variety of processors, some of which are one HELL of a lot better designed than the x86. The MIPS is a textbook example (literally) of a RISC processor, which lends itself perfectly to heavy pipelining and the like, thus making a 500 MHz MIPS a lot more impressive than it sounds. Also, the particular MIPS flavor the Chinese are working on is 64-bits, which gives them another means to approach the performance of the western chips in at least some applications without quite making the same clock speed.
I have done considerable mathematical editing in WordPerfect 5.1 and 6.0 for DOS. It was more bearable than the Microsoft Equation Editor, which I later *tried* to use, since there was some sort of TeX-like notation you could use. It got really, really messy when I tried to typeset a term paper on head-driven phrase structure grammar (an interesting branch of theoretical linguistics originated at Stanford). These days I use (La)TeX, whether there are equations in my document or not. Like emacs, or even WP 5.1, or for that matter just about any computer program that actually *increases* my productivity, it took a while to learn, and I still quite frequently refer to the Not So Short Guide and the Gentle Introduction, but it really pays off, because it's a well-designed system that wasn't shipped out in a rush by a software company that thinks ease of learning equals ease of use and employs programmers who have never opened a CS textbook.
Ho, ho, slow down... Music is a very pluriform sort of thing, that comes in many different types. Some people maybe into musical traditions that do not, traditionally, view standard music notation as an absolute requirement, such as pop, folk, or rock. Often, the use of improvisation and a more fluid means of adapting each other's work lend great charm to these genres, and I would be the last to condemn a good bit of folk, rock, blues or what-have-you.
At the same time, plenty of people immensely enjoy types of music where notation is very important. I, for one, am a great admirer and mediocre practitioner of western classical music, a particularly rich musical tradition that dates back to the middle ages. Within that particular tradition, musical notation still plays a very important role. Some of the reasons are that:
- it removes the need to memorize lengthy and complicated pieces before one can play them at all (although good musicians admittedly have a tendency to memorize anyhow)
- it can serve as an abstract medium of expression for a composer that maybe doesn't even play the instrument(s) s/he's writing for
- it can help further the ideal of playing music "as the composer intended", which has traditionally been valued highly, and only more so recently, by allowing a separation of the music from any particular performance
- it provides a medium for the theoretical study of musical patterns
- it provides a cheap and accessible way to catalogue and refer to the immense body of music composed over the centuries
- symphonic and choral pieces require immense amounts of coordination between large groups of musicians, which would be pretty much undoable without some form of notation
&c.
The fact that he's going to be charged in Australia anyway, which may very well be true, is not that relevant. The premise behind your excuse is that his situation wouldn't be any different on either side of the pacific. I would beg to differ. In the US, prison sentences are often longer than in other places, and prison conditions are atrocious to the point where violence, rape and murder between inmates are barely considered abnormal anymore. Trials are very heavily influenced by the amount of money you can spend on a lawyer. I have no particular knowledge of Australia, but I would imagine the guy's situation could be considerably different if he were tried at home.
What exactly do you mean by "updated for modern functional languages"? The main functional languages these days are Common Lisp, Scheme, OCaml, SML, and Haskell.
The original LISP machines ran Zetalisp, which had a lot of Common Lisp in it already, but still had quirks like dynamic scope. Most Lisp Machines derived from the MIT lineage (Symbolics, LMI and TI) supported Common Lisp as soon as it came around; Xerox had a competing system that ran Interlisp.
Now, if by "modern" functional language you mean ANSI Common Lisp or Scheme, I agree: it would be really awesome to have a modern Lisp Machine around, hopefully a little more RISC-like, but with the same basic philosophy. If you mean some ML dialect, though, the use of special-purpose architecture seems limited to me. The main win of the Lisp machines was that they could deal with typed values; in ML you don't need those, since types are dealt with statically.
Lastly, let me just shamelessly promote my Lisp Machine webpage:
http://www.its.caltech.edu/~weel/lispm.shtml
Given proper tail recursion, goto's are completely equivalent to function calls. Look at the following example, written in an imaginary C-like language start: twiddle(); if (...) {goto f; } else {goto g;} f: tweedle(); g: twaddle(); goto f Now with function calls (I'm going to use Scheme syntax here, but don't be scared): (define (start) (twiddle) (if ... (f) (g)))
(define (f)
(tweedle)
(g))
(define (g)
(twaddle)
(f))
You may *think* that every function call is (a) slow and (b) pushes stack, but that's not necessarily the case. The standard paper (good read!) is ftp://publications.ai.mit.edu/ai-publications/pdf/ AIM-443.pdf .
The big difference between OCaml and Lisp/Scheme is that even though both are *strongly* typed, that is, you can never, say, reference memory address pi becaus a float is simply not a pointer, they work it in different ways. Lisp/Scheme is dynamically typed, like Python, and stores the types with the values, which are checked at runtime. OCaml is statically typed, which means it does not need to store types in tags, but analyzes your program at compile time to ensure "type safety".
On the one hand, this gives a performance win. On the other hand, it makes programming slightly more painful at times. The expression "if x=3 then 5 else 2.0" is not acceptable to the OCaml compiler, because it cannot determine its type; the expression (if (eq? x 3) 5 2.0) is perfectly acceptable to a Scheme compiler. The advantage of this, it is claimed, is that most programs without well-defined types are buggy anyhow, so it helps you write good code.
Note that the same performance win can be had from not having the language be very precise about types at all, and randomly allowing dereferencing of integers and multiplication of pointers. The archetype for that, of course, is C, which is considered *weakly* typed (a.k.a. "wrongly typed"). This has the disadvantage that programs can behave unpredictably and cause system crashes very easily. If an OCaml or Scheme program ever segfaults, let alone causes a bus error, that's a bug in the compiler. If a C program does that, well you all know what that means...
Just to confuse you a little more, there's yet another distinction to be made, which is between explicitly and implicitly typed. A language such as C or Java is explicitly typed, meaning that you have to declare variables and functions as being of a certain type, although the explicit types in C are mostly sugar and can always be overridden. Python and Scheme are implicitly typed, that is, you don't mention any types, which is not a big deal, since there really is only one type. All right, backtrack. I just said they had type tags, and now I say there's only one type? The answer to that is one of terminology. From a theoretical perspective, if any slot can contain any value, there's only one type (like Java's Object).
Now OCaml is statically, strongly *and* implicitly typed. That means that it will check at compile time that there can be no type errors, and you don't even have to mention the types. Say you write a funtion in OCaml like this: let f x = 2 * x . Then it will derive for you that it has type Int -> Int. This is accomplished through a piece of higher math called Hindley-Milner type inferencing.
If you are interested in this sort of stuff, I strongly recommend Lambda the Ultimate (lambda.weblogs.com), which is always full of programming language reading material, both on a very abstruse theoretical and on a very practical level.
I beg to differ. The average Mac is in use longer than the average PC. I can only guess at the reasons.
That is definitely not the case in the US. If anyone wants to know my grades, they have to ask me, and then I can request a transcript to be sent to them by the Institute. Exceptions are made for Institute officials, but I had to explicitly sign a form authorizing them to release them to scholarship organizations. From the time I spent at UC Berkeley, I seem to remember the same is true for public universities.