Right, but it's your personal machine. They can just ban the whole thing then. If it's a public machine (like your univ. system) they can't ban the whole machine without screwing other users, but they can ban your username.
Self-serve. This is an era of vending machines and ATMs. Why revert to old tech when we already have better? Sure, refill the machine and occasional maintenance, but let's be honest -- is it cheaper to have a Coke machine or a minimum-wage guy standing around selling Cokes?
Microsoft relies on a EULA instead of straight copyright to protect their fonts, which are the relevant ones here.
And few cloned fonts have the same spacing/characters as the fonts they're trying to clone. Actually, most of MS's fonts are "clone fonts", but they differ from the Adobe originals significantly.
The fact that he has spotters outside to dig him out if he gets trapped and that he's in a prebuilt shed to give him a rather sizeable air pocket with someone else.
Take what he said at face value -- that's "sit next to". MS's people aren't (as far as I know) trying to sleep with Sen. Coverdell, and they asked for the same thing.
At some point, anti-pedophilia went hyperactive and got more than a little out of control...in the 70s the one thing that everyone could agree on bashing was Communism. Now it's pedophilia. A long, long time ago the accepted thing to bash was immigrants that came in and out-competed US workers for work -- xenophobic anti-Irish and anti-Italian sentiments.
Stretch way back, and it's Satanism and we hanged people. Why? Because you've got a majority of people who can comfortably attack something together, even if they go well beyond the bounds of reason.
Defend a communist in the 70s and you could be branded a communist yourself.
I wonder what the "thing to bash" will be in twenty years...history says that we'll find something.
If you use glib (that blessed library that makes C programming pleasant instead of miserable, not to be confused with glibc), you get a couple of extremely handy functions:
g_strconcat() (mallocs a new string that's just large enough, concatenates all passed in strings to it)
g_strdup_printf() (mallocs a new string that holds the result of a sprintf()).
Most people's exposure to glib is usually through gtk+, which uses it heavily.
The point of K&R was to jam programs onto paper books. Using K&R is to a computer what QWERTY is. It's utterly stupid to use K&R for actual programs on a computer -- BSD style is clearly the way to go.:-)
The problem is all this reading and parsing of ten zillion text-based config files. Why does JavaScript need to run to parse the prefs file? Rrrrggghhh...
Used to be a common misconception, though. Try here.
My thoughts exactly -- Exult == good
on
Ultima 7 in Windows?
·
· Score: 4, Interesting
Exult is a hell of a lot more technically *good* than this thingie, has complete Linux support (as well as Windows), and even adds a few features. If you're an Ultima fan, check it out.
Ultima, Star Control 2, Marathon...eventually, *everything* comes to Linux.
competitive with chips more than four times the clock speed
That's ridiculous. Apple has stretched this way too far, and there's a lot of misconceptions out there.
First, AltiVec is *not* going to buy you "more than four times" your base performance, no way, no how. In theory, ignoring memory bandwidth issues, the fact that you aren't going to have a string of vector instructions without any jump instructions, and the fact that no real world problem uses anything like this, AltiVec peaks out at nine times normal performance, assuming you need to do a bunch of low-precision floating point operations *and* simultaneous permutes. Anyone who has worked on a compiler knows very well that real world use of SIMD, even under best-case conditions, is nowhere *near* peak conditions, even under ideally suited problems. The best benchmarks I've seen show that on those few problems that AltiVec is useful for *and* the programmer hand-coded and tweaked the relevant inner loop, you can get perhaps a 50% performance improvement.
Motorola's been sucking when it comes to fabbign new chips, and Apple's marketing folks just lean more and more heavily on "AltiVec". It's a dream, folks.
Furthermore, it doesn't really matter much any more. Both Intel and AMD chips have their own, excellent, SIMD implementations these days (MMX, SSE, SSE2, 3dNow!), and can pull the same performance gains on parallelizable problems.
I like the PPC's architecture -- I've dicked around with assembly for both it and x86, and it's a lot cleaner. I like the sane power usage on the PPC. But no way in hell is it remotely competitive in simple performance terms with Intel's or AMD's offerings, and claims that it is are false and silly.
This is not even remotely done because of security issues.
It's pretty blatantly obvious to anyone involved the security area that security fixes that require "securing the rest of the Internet" just aren't going to work. A good example of this is the attempt to "secure the Internet against spam." The current approach -- trusting other servers on the Internet and trying to simply secure all legtimate mail servers from spammers does not work. Keep in mind that anti-spam measures have nearly universal support, a tremendous number of volunteers, high visibility, and is a well-understood problem. It's pretty well understood now that trying to secure the Internet by securing every possible point of entry is not in the least feasible. The closest anyone has come is USENET, which is a much less critical, more tightly controlled system with the Usenet Death Penalty for offending ISPs -- and even so, as USENET aficionados know, there's still a huge amount of spam.
If the OHS is scared that they won't be able to trace someone because they're coming in from a wireless port, they need to secure all the services that they're concerned about and require a digital identification of some sort. Trying to make the Internet watertight is not, no way, no how going to happen. You can't secure the US and lock the rest of the world out, and you can't secure the entire world. You can't even reasonably secure all the possible points of entry in a state.
This isn't about security. It isn't even about technology.
Ever since Bush signalled that he was willing to back just about anything that "fought terrorism", every stupid agenda out there has managed to include "fighting terrorism". People competing with 802.11b (*cough* telecom corps pushing 3G services, currently being pretty much ignored in favor of the faster, cheaper 802.11b) would love nothing better than to hand their favorite politician a few dollars to "crack down on terrorism" on 802.11b. In contrast, *their* networks are easily monitored, and as evidenced by cells in the past, telecom corps are more than happy to use key escrow and provide information to federal agents. It's a ploy to try to save all those dollars invested in 3G, the marvellous moneymaker where telecom corps can charge you by the kilobyte. It's not a security issue.
Friends, this is US politics at its best -- "campaign contributions" (bribery) at full throttle.
You know, I'm a tremendous fan of the PPC architecture versus the x86. Constant length instructions are a godsend when hacking around in existing binaries. The number of registers is wonderful, and the instruction set doesn't have the massive amount of *crap*.
However, Apple has billed AltiVec as way, *way* more useful than it actually is. SIMD is simply not useful for the overwhelming majority of code, and it's a pain to use, requiring the developer to go out of his way.
Because of the superiority of Altivec, I'm not really worried about the 970 lagging behind Intel or AMD chips.
No. AltiVec is not a bad thing, but it simply is not a substitude for raw clock speed, either.
I think that your best point is the power usage. AMD chips currently draw something like 60 watts, Intel 70. I haven't looked at PPC chips for a bit, but I believe the 1Ghz chips draw around 20 watts (actually, quite an increase -- the PPC line used to be around 5 watts).
CPU power just doesn't matter all that much any more for most people -- it really has outstripped the ability of software to usefully use it. That will, I'm sure, change, but at the moment, a 600Mhz x86 processor packs all the CPU power that anyone's going to need for most tasks. Use Linux, and it's even less. Power usage *is* an issue, as fans start producing more and more noise, and hard drives and CPUs have gotten hotter. The solution is not to try to swathe your computer in blankets of sound-insulating crap after adding scads of failure-prone fans -- it's to use components that draw less power and produce less waste heat. This is one thing that Apple has done right -- I cannot buy a reasonably-powered x86 processor that draws a sane amount of heat (sane being sub-30 watts). (Transmeta's stuff is nice for laptops, but their processors are actually a bit slow for new machines, desktop machines).
The other point I have is that reducing "average" heat production is a good thing, but not a substitute for peak heat production. My system *is* going to have to undergo peak heat production at some times -- the hard drive seeking, the CPU running at 100% -- and people need to look at the numbers for *that*, not "average heat".
Slashdot Mirror
I think the point is that the boycott failed -- amazon didn't pay any attention, and most consumers ignored it.
feh /proc?)
gtk-gnutella
lopster
dc_gui w/ dctc
mtr
gkrellm (Not sure how well this would work...do OS X systems have a compatible
xmms
Embarrassingly enough, for once the "SOVIET RUSSIA" post was funnier than the +5 Funny original post...
Right, but it's your personal machine. They can just ban the whole thing then. If it's a public machine (like your univ. system) they can't ban the whole machine without screwing other users, but they can ban your username.
Self-serve. This is an era of vending machines and ATMs. Why revert to old tech when we already have better? Sure, refill the machine and occasional maintenance, but let's be honest -- is it cheaper to have a Coke machine or a minimum-wage guy standing around selling Cokes?
Microsoft relies on a EULA instead of straight copyright to protect their fonts, which are the relevant ones here.
And few cloned fonts have the same spacing/characters as the fonts they're trying to clone. Actually, most of MS's fonts are "clone fonts", but they differ from the Adobe originals significantly.
The fact that he has spotters outside to dig him out if he gets trapped and that he's in a prebuilt shed to give him a rather sizeable air pocket with someone else.
Take what he said at face value -- that's "sit next to". MS's people aren't (as far as I know) trying to sleep with Sen. Coverdell, and they asked for the same thing.
At some point, anti-pedophilia went hyperactive and got more than a little out of control...in the 70s the one thing that everyone could agree on bashing was Communism. Now it's pedophilia. A long, long time ago the accepted thing to bash was immigrants that came in and out-competed US workers for work -- xenophobic anti-Irish and anti-Italian sentiments.
Stretch way back, and it's Satanism and we hanged people. Why? Because you've got a majority of people who can comfortably attack something together, even if they go well beyond the bounds of reason.
Defend a communist in the 70s and you could be branded a communist yourself.
I wonder what the "thing to bash" will be in twenty years...history says that we'll find something.
with an HTML enabled mail reader, sometimes the first thing they saw was some pornographic picture.
The obvious solution would be to not use an HTML "enabled" mail reader...
TrueType is a propriatary font format?
The format isn't, but Times New Roman and friends are.
Wee bit of a functional bias, eh?
You're right that C limits large-scale optimization too much, though...
If you use glib (that blessed library that makes C programming pleasant instead of miserable, not to be confused with glibc), you get a couple of extremely handy functions:
g_strconcat() (mallocs a new string that's just large enough, concatenates all passed in strings to it)
g_strdup_printf() (mallocs a new string that holds the result of a sprintf()).
Most people's exposure to glib is usually through gtk+, which uses it heavily.
The point of K&R was to jam programs onto paper books. Using K&R is to a computer what QWERTY is. It's utterly stupid to use K&R for actual programs on a computer -- BSD style is clearly the way to go. :-)
"First in the morning I wrote Secure, Efficient and Easy C Programming Mini-HOWTO..."
Damn. What are your plans for the rest of the day?
"If you've done six impossible things this morning, why not round it off with breakfast at Milliways, the Restauraunt at the End of the Universe?"
-- Douglas Adams
The problem is all this reading and parsing of ten zillion text-based config files. Why does JavaScript need to run to parse the prefs file? Rrrrggghhh...
Internet Explorer is an integrated part of the desktop, therefore it is inherently faster than any other programs.
*snort* snicker chuckle
Ahh....
Used to be a common misconception, though. Try here.
Exult is a hell of a lot more technically *good* than this thingie, has complete Linux support (as well as Windows), and even adds a few features. If you're an Ultima fan, check it out.
Ultima, Star Control 2, Marathon...eventually, *everything* comes to Linux.
competitive with chips more than four times the clock speed
That's ridiculous. Apple has stretched this way too far, and there's a lot of misconceptions out there.
First, AltiVec is *not* going to buy you "more than four times" your base performance, no way, no how. In theory, ignoring memory bandwidth issues, the fact that you aren't going to have a string of vector instructions without any jump instructions, and the fact that no real world problem uses anything like this, AltiVec peaks out at nine times normal performance, assuming you need to do a bunch of low-precision floating point operations *and* simultaneous permutes. Anyone who has worked on a compiler knows very well that real world use of SIMD, even under best-case conditions, is nowhere *near* peak conditions, even under ideally suited problems. The best benchmarks I've seen show that on those few problems that AltiVec is useful for *and* the programmer hand-coded and tweaked the relevant inner loop, you can get perhaps a 50% performance improvement.
Motorola's been sucking when it comes to fabbign new chips, and Apple's marketing folks just lean more and more heavily on "AltiVec". It's a dream, folks.
Furthermore, it doesn't really matter much any more. Both Intel and AMD chips have their own, excellent, SIMD implementations these days (MMX, SSE, SSE2, 3dNow!), and can pull the same performance gains on parallelizable problems.
I like the PPC's architecture -- I've dicked around with assembly for both it and x86, and it's a lot cleaner. I like the sane power usage on the PPC. But no way in hell is it remotely competitive in simple performance terms with Intel's or AMD's offerings, and claims that it is are false and silly.
This is not even remotely done because of security issues.
It's pretty blatantly obvious to anyone involved the security area that security fixes that require "securing the rest of the Internet" just aren't going to work. A good example of this is the attempt to "secure the Internet against spam." The current approach -- trusting other servers on the Internet and trying to simply secure all legtimate mail servers from spammers does not work. Keep in mind that anti-spam measures have nearly universal support, a tremendous number of volunteers, high visibility, and is a well-understood problem. It's pretty well understood now that trying to secure the Internet by securing every possible point of entry is not in the least feasible. The closest anyone has come is USENET, which is a much less critical, more tightly controlled system with the Usenet Death Penalty for offending ISPs -- and even so, as USENET aficionados know, there's still a huge amount of spam.
If the OHS is scared that they won't be able to trace someone because they're coming in from a wireless port, they need to secure all the services that they're concerned about and require a digital identification of some sort. Trying to make the Internet watertight is not, no way, no how going to happen. You can't secure the US and lock the rest of the world out, and you can't secure the entire world. You can't even reasonably secure all the possible points of entry in a state.
This isn't about security. It isn't even about technology.
Ever since Bush signalled that he was willing to back just about anything that "fought terrorism", every stupid agenda out there has managed to include "fighting terrorism". People competing with 802.11b (*cough* telecom corps pushing 3G services, currently being pretty much ignored in favor of the faster, cheaper 802.11b) would love nothing better than to hand their favorite politician a few dollars to "crack down on terrorism" on 802.11b. In contrast, *their* networks are easily monitored, and as evidenced by cells in the past, telecom corps are more than happy to use key escrow and provide information to federal agents. It's a ploy to try to save all those dollars invested in 3G, the marvellous moneymaker where telecom corps can charge you by the kilobyte. It's not a security issue.
Friends, this is US politics at its best -- "campaign contributions" (bribery) at full throttle.
You know, I'm a tremendous fan of the PPC architecture versus the x86. Constant length instructions are a godsend when hacking around in existing binaries. The number of registers is wonderful, and the instruction set doesn't have the massive amount of *crap*.
However, Apple has billed AltiVec as way, *way* more useful than it actually is. SIMD is simply not useful for the overwhelming majority of code, and it's a pain to use, requiring the developer to go out of his way.
Because of the superiority of Altivec, I'm not really worried about the 970 lagging behind Intel or AMD chips.
No. AltiVec is not a bad thing, but it simply is not a substitude for raw clock speed, either.
I think that your best point is the power usage. AMD chips currently draw something like 60 watts, Intel 70. I haven't looked at PPC chips for a bit, but I believe the 1Ghz chips draw around 20 watts (actually, quite an increase -- the PPC line used to be around 5 watts).
CPU power just doesn't matter all that much any more for most people -- it really has outstripped the ability of software to usefully use it. That will, I'm sure, change, but at the moment, a 600Mhz x86 processor packs all the CPU power that anyone's going to need for most tasks. Use Linux, and it's even less. Power usage *is* an issue, as fans start producing more and more noise, and hard drives and CPUs have gotten hotter. The solution is not to try to swathe your computer in blankets of sound-insulating crap after adding scads of failure-prone fans -- it's to use components that draw less power and produce less waste heat. This is one thing that Apple has done right -- I cannot buy a reasonably-powered x86 processor that draws a sane amount of heat (sane being sub-30 watts). (Transmeta's stuff is nice for laptops, but their processors are actually a bit slow for new machines, desktop machines).
The other point I have is that reducing "average" heat production is a good thing, but not a substitute for peak heat production. My system *is* going to have to undergo peak heat production at some times -- the hard drive seeking, the CPU running at 100% -- and people need to look at the numbers for *that*, not "average heat".
This is very true.
If we went by the "hello world metric", Perl would probably be consider the best of all programming languages.
From what I've heard, the majority of eiffel developers do very large software applications.
If you're trying to churn out useful software and not do research and want to learn ML, you may prefer Objective Caml to SML.
I really *hate* type inference, as it makes debugging a pain, and SML/NJ, the SML I learned on, has absolutely atrocious error reporting.
On the up side, it's quite an eye opener if you've never used a language with really strong typing.
For a "modern", safe language, SmartEiffel produces very fast code. I don't know of any other imperative language other than C and C++ that beats it.
I think the issue is that it can't be that difficult to add something to the synthetic stuff to make it flame-retardant.
:-)
Of course, that might just eliminate the possibility of it being non-carciogenic.