>...type-inferencing Lisp compilers... > Sun could have... made the language dynamically typed in the first place...
You're confusing type inference with dynamic typing. Type inference is a feature of statically typed languages such as ML and Haskell, whereby the static types are inferred from the uses of variables and functions, rather than being declared by the programmer. The end result is to provide most of the convenience of dynamic typing (easy generics, concise code, quicker prototyping), while retaining the speed* and compile-time type-checking of static typing.
* Okay, in the case of Haskell "not retaining the speed" would be a more accurate way of putting it, but that's beside the point.
Unicode will remain limited to U+0.. U+10FFFF (the highest codepoint UTF-16 can represent via surrogates) for a long time, possibly forever.
Definitely forever; it's inscribed on stone tablets kept in a hidden vault deep below the secret ISO headquarters on the dark side of the moon.
UCS-4 is absurdly wasteful (at least eleven bits per character are never used) so UTF-8 and UTF-16 are both reasonable compressed encodings...
Two points:
1. UTF32 (what the Unicode consortium calls what you call UCS-4) is more efficient than the smaller encodings. Your arrays of 16-bit characters (if you're using UTF16 internally) are likely to be aligned on 32-bit boundaries anyway, so you're actually wasting space using UTF16 if you're using characters outside the BMP. Besides, for most applications, strings aren't that large anyway. Does it really matter if your data is using 100 kb instead of 60 kb?
2. UTF-8 is absurdly wasteful. Not only is it slow to encode and decode, but it also uses more space than UTF16 for any far-Eastern script.
Mixing multiple languages in a single
project is an enormous evil, because it means that all maintainers have to be fluent in both toolchains, and in the integration layer.
I'm tempted to apply reductio ad absurdum here, and in fact I think I will:
Mixing multiple tools in a single carpentry project is an enormous evil, because it means that all carpenters have to know how to use planes as well as saws.
Let's face it, different languages DO have different strengths and different uses. That's why Fortran is still popular in the scientific community.
In fact, mixing multiple languages in a single project has been the standard practice ever since the first high-level languages appeared. We might no longer consider dropping as far as ASM, but it still makes a lot of sense to write speed-critical code in C or C++, and the main application framework in something higher-level.
> So why the hell would Microsoft actively REMOVE Linux support when it was already a feature / part of the program?
They haven't "removed Linux support": it's MICROSOFT that won't support Linux, not the program. What that means is, if you're having trouble running Linux, Microsoft won't help you get it to work. That sounds reasonable to me.
> VMWARE? Perhaps Bochs should be considered by all of us as the free alternative.
I'm sorry, but comparing Bochs in its current state to Virtual PC (I have the old Connectix version) is a bit like comparing DOS on an 8086 to Windows XP on a modern PC. In fact, that's almost a perfect comparison, since I've only ever managed to get Bochs to simulate an 8086 running DOS. Which it does very accurately.
> I couldn't find crap on the masterlock site (who on earth can navigate that site?)
I clicked on his link, then I clicked on the bit where it said "Customer Service", then I clicked on the bit in the menu that appeared that said "Anti-theft guarantee". In other words, I successfully navigated to the relevant page in under ten seconds without any false starts.
If you want to bitch about hard-to-navigate sites, there are plenty of hard-to-navigate sites out there to choose from without picking on relatively harmless ones like this.
The relevant information, by the way, is that all Masterlock locks come with a limited guarantee to the effect that if your bike is stolen, while locked with their lock, and the thief achieved this by breaking or forcing the lock, then you'll get some money if you manage to jump through a litany of hoops.
> Can you say Win98? Nothing more than Win95 with an expanded filesystem and a few rinky dink addons? That should have been a simple free download not a much trumpeted "Upgrade".
Wrong. Win98 was a vast improvement on Win95.
You're succumbing to the usual error of MS-bashers, which is that you attack the wrong things. If you want to criticise MS's "upgrades", for example, may I suggest attacking the shifts from Win98 to Win98 SE, or from Win98 SE to WinMe? *Those* are the ones that should have been free downloads.
You sound like someone who's just bought an expensive new car, and is complaining that it doesn't clean their carpets very well.
Surely the whole purpose of this is to find a book which discusses some particular subject? That's not USEFUL for fiction. Try it with, I don't know, the names of authors, rather than their characters, and you'll probably get some more useful results from critical works about those authors. Try it with the names of scientific processes, perhaps. Try it on algorithms. And if you want to find quotations, try a dictionary of quotations. Horses for courses, and all that.
Okay, and then the next version might display all text using CAPTCHA-style distortion to baffle OCR software. Wouldn't stop you typing it all in again, but the idea here is to make copying this stuff impractical, not impossible.
> I wonder if the British bookstores buy books back and resell them in later semesters?
Yes - but not as stingily as it appears American ones do. Most of the comments here mention American bookstores paying $10 for a book that cost $120 originally, then selling the used copy for $100. My experience in the UK was bookstores promising to pay back half the price you paid (on production of a receipt), and then reselling the books for only a little more than that (say 65 for a 100 book).
> At least until he's trumped by the powers of communism (lawsuits by the school or the textbook becoming illegal to import under the DMCA)
Um, what?
Exploitative monopolies are NOT communism. Lawsuits or legislation against importing a textbook would be symptomatic of the corporatism that the USA seems hell-bent on forcing down the world's throat.
The communist scenario would see the textbooks provided to everyone for free. Of course, once students graduated they would be expected to work for free as well, but you can't have everything.
Here's why the books are cheaper in other countries:
If they weren't the US companies with overseas sweatshop professional workers would have to pay them higher salaries so they could afford to go to school.
Um, I hate to break it to you, but there aren't many sweatshops in the UK filled with workers who can't afford to go to school. We're, like, a first-world country?
> Second of all, are you saying the Drake Equation is crap? That there's simply nothing out there?
The former does not imply the latter. The Drake Equation *is* crap, by which I mean it has absolutely no scientific value whatsoever. It's a fancy way of saying "we don't know".
Fortunately SETI does have some scientific methodology behind it as well.
> It will be as easy as getting the US to... transition with the rest of the world to driving on the left side of the road.
I was under the impression that the side of the road on which people drove was split fairly evenly between left / right? Britain, parts of the Commonwealth, and Japan on the left, the USA and mainland Europe on the right, and India straight down the middle except when there's a cow in the way.
Slashdot Mirror
> ...type-inferencing Lisp compilers... ... made the language dynamically typed in the first place...
> Sun could have
You're confusing type inference with dynamic typing. Type inference is a feature of statically typed languages such as ML and Haskell, whereby the static types are inferred from the uses of variables and functions, rather than being declared by the programmer. The end result is to provide most of the convenience of dynamic typing (easy generics, concise code, quicker prototyping), while retaining the speed* and compile-time type-checking of static typing.
* Okay, in the case of Haskell "not retaining the speed" would be a more accurate way of putting it, but that's beside the point.
Two points:
1. UTF32 (what the Unicode consortium calls what you call UCS-4) is more efficient than the smaller encodings. Your arrays of 16-bit characters (if you're using UTF16 internally) are likely to be aligned on 32-bit boundaries anyway, so you're actually wasting space using UTF16 if you're using characters outside the BMP. Besides, for most applications, strings aren't that large anyway. Does it really matter if your data is using 100 kb instead of 60 kb?
2. UTF-8 is absurdly wasteful. Not only is it slow to encode and decode, but it also uses more space than UTF16 for any far-Eastern script.
See also the Unicode FAQ on UTFs.
Mixing multiple tools in a single carpentry project is an enormous evil, because it means that all carpenters have to know how to use planes as well as saws.
Let's face it, different languages DO have different strengths and different uses. That's why Fortran is still popular in the scientific community.
In fact, mixing multiple languages in a single project has been the standard practice ever since the first high-level languages appeared. We might no longer consider dropping as far as ASM, but it still makes a lot of sense to write speed-critical code in C or C++, and the main application framework in something higher-level.
> So why the hell would Microsoft actively REMOVE Linux support when it was already a feature / part of the program?
They haven't "removed Linux support": it's MICROSOFT that won't support Linux, not the program. What that means is, if you're having trouble running Linux, Microsoft won't help you get it to work. That sounds reasonable to me.
Linux still works just as well as it did before.
Move along, nothing to see here...
> VMWARE? Perhaps Bochs should be considered by all of us as the free alternative.
I'm sorry, but comparing Bochs in its current state to Virtual PC (I have the old Connectix version) is a bit like comparing DOS on an 8086 to Windows XP on a modern PC. In fact, that's almost a perfect comparison, since I've only ever managed to get Bochs to simulate an 8086 running DOS. Which it does very accurately.
The problem is with the website. Decreasing your text size (View/Decrease Text Size) should enable you to read the full text.
In other words, the person you should be filing the bug report with is the webmaster at self-confidence.co.uk.
> Too bad that whole 'football' thing never caught on.
Unfortunately it did - it's that funny thing they play in America that the world hasn't heard of.
I suggest "Beckham Mozilla" for outside the USA.
> I couldn't find crap on the masterlock site (who on earth can navigate that site?)
I clicked on his link, then I clicked on the bit where it said "Customer Service", then I clicked on the bit in the menu that appeared that said "Anti-theft guarantee". In other words, I successfully navigated to the relevant page in under ten seconds without any false starts.
If you want to bitch about hard-to-navigate sites, there are plenty of hard-to-navigate sites out there to choose from without picking on relatively harmless ones like this.
The relevant information, by the way, is that all Masterlock locks come with a limited guarantee to the effect that if your bike is stolen, while locked with their lock, and the thief achieved this by breaking or forcing the lock, then you'll get some money if you manage to jump through a litany of hoops.
You mean the next version of Windows will be written in Haskell? Excellent, that should eliminate the buffer overflows and speed things up a bit.
> Can you say Win98? Nothing more than Win95 with an expanded filesystem and a few rinky dink addons? That should have been a simple free download not a much trumpeted "Upgrade".
Wrong. Win98 was a vast improvement on Win95.
You're succumbing to the usual error of MS-bashers, which is that you attack the wrong things. If you want to criticise MS's "upgrades", for example, may I suggest attacking the shifts from Win98 to Win98 SE, or from Win98 SE to WinMe? *Those* are the ones that should have been free downloads.
You sound like someone who's just bought an expensive new car, and is complaining that it doesn't clean their carpets very well.
Surely the whole purpose of this is to find a book which discusses some particular subject? That's not USEFUL for fiction. Try it with, I don't know, the names of authors, rather than their characters, and you'll probably get some more useful results from critical works about those authors. Try it with the names of scientific processes, perhaps. Try it on algorithms. And if you want to find quotations, try a dictionary of quotations. Horses for courses, and all that.
> Ever hear of Optical Character Recognition?
Okay, and then the next version might display all text using CAPTCHA-style distortion to baffle OCR software. Wouldn't stop you typing it all in again, but the idea here is to make copying this stuff impractical, not impossible.
> ...expanding acronyms and including a URL or two...
Maybe your point would have been better made if you'd said "link"?
> I wonder if the British bookstores buy books back and resell them in later semesters?
Yes - but not as stingily as it appears American ones do. Most of the comments here mention American bookstores paying $10 for a book that cost $120 originally, then selling the used copy for $100. My experience in the UK was bookstores promising to pay back half the price you paid (on production of a receipt), and then reselling the books for only a little more than that (say 65 for a 100 book).
> At least until he's trumped by the powers of communism (lawsuits by the school or the textbook becoming illegal to import under the DMCA)
Um, what?
Exploitative monopolies are NOT communism. Lawsuits or legislation against importing a textbook would be symptomatic of the corporatism that the USA seems hell-bent on forcing down the world's throat.
The communist scenario would see the textbooks provided to everyone for free. Of course, once students graduated they would be expected to work for free as well, but you can't have everything.
Um, I hate to break it to you, but there aren't many sweatshops in the UK filled with workers who can't afford to go to school. We're, like, a first-world country?
> Either that, or run for president and divulge all this information to the public.
I'd love to. Really, I'd love to. Unfortunately I'm an alien.
(...in the sense of "not a US citizen", I hasten to add!)
> Second of all, are you saying the Drake Equation is crap? That there's simply nothing out there?
The former does not imply the latter. The Drake Equation *is* crap, by which I mean it has absolutely no scientific value whatsoever. It's a fancy way of saying "we don't know".
Fortunately SETI does have some scientific methodology behind it as well.
> They knew English and one of them said he would have recognized Russian.
I'm glad we have the testimony of an expert linguist to rely on.
> Alien species would be so alien we wouldn't be able to communicate with them without considerable intelligence and effort.
What, you mean like lawyers?
If all you want is his credit card number, wouldn't it be cheaper to set up a fake pr0n site?
> It will be as easy as getting the US to ... transition with the rest of the world to driving on the left side of the road.
I was under the impression that the side of the road on which people drove was split fairly evenly between left / right? Britain, parts of the Commonwealth, and Japan on the left, the USA and mainland Europe on the right, and India straight down the middle except when there's a cow in the way.
> Rouge Squadron
Cue predictable jokes about makeup.