I thought this stupid thread would've died already, but since it hasn't let's use some high school physics to show how wrong you are.
*Suppose* that 100 million years ago the earth's day were only 18 hours long. I don't know if it was, but suppose that.
Then the measured gravity acceleration would be
g = g_0 - Rw^2, where w is omega (the earth's period)
w = 2pi/64800 g = 9.8 - 6,37e6*(9.7e-5)^2 = 9.8 - 0.06 = 9.74 m/s^2
So I can't see how g could've been about 15.2 m/s^2, because reducing earth's period doesn't make much of a difference (as many people have stated without proof before me).
What amazes me is that you state that g was actually HIGHER (15.2 m/s^2) back in that day. Would you mind elucidating that?
> Now, if you just took a simple program like sorting, well, hell, C++ blows Java away. But if you have a complex program that runs for a good while, then Java matches or even beats out C++ and even C. It's all about the cache.
Can you direct me to (non-trivial) code that demonstrates this? I'd like to experiment with it, because it seems really neat.
> There are many assemblers out there written in C, C++ and Pascal yet the production language is faster. What is your point ?
First, an assembler and a VM are two very different programs. An assembler can be written in BASIC and produce faster code than an assembler written in C. A VM written in BASIC will NOT run your bytecode faster than a VM written in C.
What's _your_ point? You've just shown to be ignorant regarding a very simple concept.
Second, I never said GC didn't have any benefits and I specially don't need you to give me a link to a primer to GC.
I simply claimed there's no way a language that's compiled into bytecode and has GC will be as fast as regular C++ as it's claimed in the article. Don't turn this into a Java vs. C++ contest and specially think before you post.
We had a chance to sit down and chat with the man in charge of this initiative, Chris Melissinos, chief gaming officer for Sun Microsystems, as he was preparing for his GDC presentation.
That doesn't mean he's a marketroid. If he is, we shouldn't be wasting our time with him.
"There are a lot of misconceptions about Java," explains Melissinos, "one is that it's too slow".
Ok, let's suppose for a minute that Java is _not_ slow and that this guy's correct. Then we see the following:
Game developers would say, 'You'll never get C code to run as fast as Assembly, you'll never do it.' Well it happened. When C++ came out, the same thing occurred, and once again C++ became the development standard. The same thing holds true today, folks are saying there is no way Java is as fast as C++. Well I'm here to show you it can and it can even run faster."
I find it impossible to believe that a bytecode program, which runs on top of a VM (written, by the way, in C++ or in an equivalent imperative language) could be faster than C++ unless the programmer is absolutely clueless. But I'll even be satisfied with "only 50% slower" if I see an example.
What I won't take is reading the previous two and the following in the same damn article:
"Developers are saying that it's so much easier to code in Java than in C++," says Melissinos, "and there is less to worry about as well. The Java code actually takes care of problems like garbage collection and memory issues that C++ doesn't. In C++ you have to do all that manually and Java does it automatically saving so much time in development."
So we basically have this guy saying that Java can be faster than C++, which can be as fast as assembly and still admitting that Java uses garbage collection (an inherently slow task). He's either so passionate about Java that he doesn't realise what he's saying or he's been spending too much time with the people in marketing.
This is called a SIMD (single instruction, multiple data) operation. It's what MMX is all about.
It's usually not worth doing this if there's no SIMD hardware support, because the time wasted loading your values and then separating them isn't compensated by the gain in speed. Of course there are special cases (like when dealing with bit strings) where this is used by definition (and will be an improvement).
Of course it's more responsive and it just proves my point.
If you noticed a difference by adding RAM, it means the old amount wasn't enough. If XP boots faster, then this means it either uses outrageous amounts of RAM for itself (and starts swapping even while booting!) OR that it has horrible memory management (meaning it takes too long to reclaim memory that has already been freed and/or that's being used as buffers/cache).
Something that isn't pointed out enough
on
SuSE 7.3 vs XP
·
· Score: 5, Interesting
I can't read the article (it's slashdotted), but there's something about Windows (specially XP) that I rarely hear people talk about: it uses outrageous amounts of RAM. Yes, RAM is cheap, but I find it extremely suspicious when simple applications consume so many resources.
I have 128 MB of RAM and with Linux it's enough for everything I need, _including_ Mozilla (which as we all know, can use a lot of memory).
I find it ludicrous that I can't even boot XP without swapping and it takes forever to open up apps like the media player. Should I face this with a smile and say "well, I'm at fault because 128 MB of RAM clearly isn't enough"?
I can't bring myself to respect an OS which needs this many resources to do nothing. Yet I know people with 64 MB of RAM who praise XP in favor of Linux. I firmly believe that they either don't use their computers for anything productive or they lie.
... was to install Spambouncer, which is a large set of procmail filters.
Before installing it, I got ~20 spam messages a day. Now I get at most 1-2 a week. Spambouncer does come with very restrictive default settings, though. For example, you must specify if you want to receive email from free web mail services like Yahoo and Hotmail, otherwise it'll filter those out.
It also logs everything it does and has the option of sending blocked email to a file instead of/dev/null in the case it filters something it shouldn't.
In my case the only inconvenience was it blocked legitimate email from Amazon.com and eBay -- these are filled with disclaimers and have HTML, which Spambouncer doesn't like to see. In any case, it's easy to mark those domains as safe and start receiving their email again.
So we know what their secret to running windows apps is: Wine.
What I find unsettling is that this sneak preview states 99% of Windows applications go "poof" on install (which is not very surprising for anyone who has ever used wine). Yet we've seen screenshots showing Internet Explorer on top of KDE.
So another question stands: are those screenshots real?
Maybe YOUR company won't consider you if you don't have a BS. There are plenty of companies who will give you a chance, however.
With lower salaries, unless you can really prove yourself. I doubt the Ask Slashdot guy will be able to do that since he's so young.
A college degree is better of course, but you're really wrong about needing one.
I wouldn't like to be out of work, 20 years old, with 5 years of questionable industry experience and WITHOUT a college degree.
You're just saying that because of the following:
As you move along in your career a degree matters less and less however. If you are 30 and have proven yourself with a bunch of good years of experience behind you, 95% of 'professionals' will actually respect you, whether or not you got that piece of paper 10 years ago.
You're correct that as you prove yourself the degree becomes less important. Until you've done that, however, you may live through some unpleasant times. If I were 20 years old and in his place, I'd try to go to school. It's much more pleasant than going against the stablished order and gives you better odds.
You have a good point, but I don't believe college is overrated.
I agree with you that college won't turn a "commoner" into a computer wizard and you as an employer are usually looking for the wizard types. The IT market usually looks for technical skills that are usually NOT taught in college. That's because one can learn those on his own. It's much more difficult to learn math on your own than it is with a teacher and that's why colleges may choose to teach math.
College may be overrated in the sense tons of college graduates (the ones who haven't got "it") go out unprepared into the real world. But that doesn't mean we should all ditch college! The unprepared ones can pick up the technical stuff they lack as they work in the real world. They'll still retain the theory they learnt, which will eventually help them in the future. Of course they still won't have "it", but that's no surprise. Most people (CS graduates or not) don't.
We'd both rather have the gifted employee who doesn't need college because he's brilliant. However, I'd rather have the brilliant type who went to college than the brilliant type who didn't. College gives you background which allows you to make better decisions because you're informed. No matter how much you've got "it", you won't be able to figure out the last 100 years of computer science without learning it from books. I believe college is the best place to do your learning.
I agree with you that having "it" is more important than anything else. That's what makes a good CS professional. What I wrote in the original post is advice to the Ask Slashdot guy. Not everybody out there agrees with you and me, and I believe he'll be much more stable with a degree.
[In short: he may be in some trouble if gets fired and he doesn't have a degree. That's why I recommend him to get one while he can. I wish we could all open our little software shops and code away without ever needing to prove ourselves with certificates of learning, but most of us can't.]
It seems to me like this is the "Hi, I'm 19-24, never went to college, am a kick ass programmer with TONS of experience but am stuck professionally! HELP!!!!" week.
I don't intend to offend you, but you must know some things:
1. Face it: college is IMPORTANT. Go study now while you still can. You may argue that you're married with kids now, but you're still young -- don't wait until you're 30 to say "I should've gone to college back when I was 20 and could".
2. Programming is cool, but it's not everything that matters. Once you learn calculus, statistics and logic you'll have better ways to solve problems. You'll see that there are solutions other than brute force.
3. Programming can be extremely easy next to mathematics. It's often much simpler to devise something that "just works" rather than developing it carefully and proving why it's the best solution.
4. Cutting and pasting perl/php for 5 years doesn't count as real work. Nor does assembling computers/networks. Just installing and updating software doesn't count as professional system administration.
5. The computer stuff you did in high school doesn't count as real world experience.
6. Whether you like it or not, most companies will NEVER consider you and most professionals won't respect you if you don't have a degree. You will keep losing arguments even though you're right.
7. You'll never know how much you lack unless you go to college.
In short, you may be stuck professionally for a reason other than age discrimination. Perhaps being a kick ass programmer isn't all that matters.
Bad news for UNIX is always something minor. That's justified, because the diversity among UNIX machines is higher (meaning bugs rarely affect a lot of people at once) and because UNIX people are more competent.
I'd say the most significant recent UNIX-related exploit was SSH's. (The one which gave free access to accounts with '*' in their password entries). And that's just UNIX-related, because SSH isn't UNIX. Furthermore, the bug doesn't affect OpenSSH, which everyone should be using anyway.
Microsoft bugs, on the other hand, pop up every month and are HUGE, like amazingly virulent worms like Code Red, IE exploits and the recent Windows XP one.
I've never used AtheOS and before giving it a try I'd like to see information regarding its hardware requirements and performance. For example, how does AtheOS perform as a web/ftp/samba/db server, considering it supports multithreading and SMP?
I don't want to be mean, but I also have a rev. A iMac with 96 MB of RAM. It runs Linux though, and it's blazing fast.
Never swaps and is extremely responsive, even when I run mozilla on it (okay, perhaps not THAT responsive in this case). And I'm also very excited about the possibilities of perl on Linux. (the dev tools also work fine.)
You're yet another one who's distorting reality, omitting facts or making proofless assertions ("the money isn't in software") to make your point logical.
I know some people who have cable connections and CD burners. The ones who know what MP3s are usually download them. Most of these people _never_ buy CDs. The same goes with software.
I'll call it a fact that people who CAN pirate WILL pirate because I have seen this much more often than not.
And the money IS in software. In absolute numbers, much more than in hardware.
Slashdot Mirror
I wasn't paying attention: w isn't the earth's period, but its angular velocity.
I thought this stupid thread would've died already, but since it hasn't let's use some high school physics to show how wrong you are.
*Suppose* that 100 million years ago the earth's day were only 18 hours long. I don't know if it was, but suppose that.
Then the measured gravity acceleration would be
g = g_0 - Rw^2, where w is omega (the earth's period)
w = 2pi/64800
g = 9.8 - 6,37e6*(9.7e-5)^2 = 9.8 - 0.06 = 9.74 m/s^2
So I can't see how g could've been about 15.2 m/s^2, because reducing earth's period doesn't make much of a difference (as many people have stated without proof before me).
What amazes me is that you state that g was actually HIGHER (15.2 m/s^2) back in that day. Would you mind elucidating that?
> Now, if you just took a simple program like sorting, well, hell, C++ blows Java away. But if you have a complex program that runs for a good while, then Java matches or even beats out C++ and even C. It's all about the cache.
Can you direct me to (non-trivial) code that demonstrates this? I'd like to experiment with it, because it seems really neat.
> There are many assemblers out there written in C, C++ and Pascal yet the production language is faster. What is your point ?
First, an assembler and a VM are two very different programs. An assembler can be written in BASIC and produce faster code than an assembler written in C. A VM written in BASIC will NOT run your bytecode faster than a VM written in C.
What's _your_ point? You've just shown to be ignorant regarding a very simple concept.
Second, I never said GC didn't have any benefits and I specially don't need you to give me a link to a primer to GC.
I simply claimed there's no way a language that's compiled into bytecode and has GC will be as fast as regular C++ as it's claimed in the article. Don't turn this into a Java vs. C++ contest and specially think before you post.
We had a chance to sit down and chat with the man in charge of this initiative, Chris Melissinos, chief gaming officer for Sun Microsystems, as he was preparing for his GDC presentation.
That doesn't mean he's a marketroid. If he is, we shouldn't be wasting our time with him.
First off,
"There are a lot of misconceptions about Java," explains Melissinos, "one is that it's too slow".
Ok, let's suppose for a minute that Java is _not_ slow and that this guy's correct. Then we see the following:
Game developers would say, 'You'll never get C code to run as fast as Assembly, you'll never do it.' Well it happened. When C++ came out, the same thing occurred, and once again C++ became the development standard. The same thing holds true today, folks are saying there is no way Java is as fast as C++. Well I'm here to show you it can and it can even run faster."
I find it impossible to believe that a bytecode program, which runs on top of a VM (written, by the way, in C++ or in an equivalent imperative language) could be faster than C++ unless the programmer is absolutely clueless. But I'll even be satisfied with "only 50% slower" if I see an example.
What I won't take is reading the previous two and the following in the same damn article:
"Developers are saying that it's so much easier to code in Java than in C++," says Melissinos, "and there is less to worry about as well. The Java code actually takes care of problems like garbage collection and memory issues that C++ doesn't. In C++ you have to do all that manually and Java does it automatically saving so much time in development."
So we basically have this guy saying that Java can be faster than C++, which can be as fast as assembly and still admitting that Java uses garbage collection (an inherently slow task). He's either so passionate about Java that he doesn't realise what he's saying or he's been spending too much time with the people in marketing.
This is called a SIMD (single instruction, multiple data) operation. It's what MMX is all about.
It's usually not worth doing this if there's no SIMD hardware support, because the time wasted loading your values and then separating them isn't compensated by the gain in speed. Of course there are special cases (like when dealing with bit strings) where this is used by definition (and will be an improvement).
Of course it's more responsive and it just proves my point.
If you noticed a difference by adding RAM, it means the old amount wasn't enough. If XP boots faster, then this means it either uses outrageous amounts of RAM for itself (and starts swapping even while booting!) OR that it has horrible memory management (meaning it takes too long to reclaim memory that has already been freed and/or that's being used as buffers/cache).
I can't read the article (it's slashdotted), but there's something about Windows (specially XP) that I rarely hear people talk about: it uses outrageous amounts of RAM. Yes, RAM is cheap, but I find it extremely suspicious when simple applications consume so many resources.
I have 128 MB of RAM and with Linux it's enough for everything I need, _including_ Mozilla (which as we all know, can use a lot of memory).
I find it ludicrous that I can't even boot XP without swapping and it takes forever to open up apps like the media player. Should I face this with a smile and say "well, I'm at fault because 128 MB of RAM clearly isn't enough"?
I can't bring myself to respect an OS which needs this many resources to do nothing. Yet I know people with 64 MB of RAM who praise XP in favor of Linux. I firmly believe that they either don't use their computers for anything productive or they lie.
... was to install Spambouncer, which is a large set of procmail filters.
/dev/null in the case it filters something it shouldn't.
Before installing it, I got ~20 spam messages a day. Now I get at most 1-2 a week. Spambouncer does come with very restrictive default settings, though. For example, you must specify if you want to receive email from free web mail services like Yahoo and Hotmail, otherwise it'll filter those out.
It also logs everything it does and has the option of sending blocked email to a file instead of
In my case the only inconvenience was it blocked legitimate email from Amazon.com and eBay -- these are filled with disclaimers and have HTML, which Spambouncer doesn't like to see. In any case, it's easy to mark those domains as safe and start receiving their email again.
Yeah, but that was only because he didn't install IE and Office from scratch -- he installed those on Windows NT first.
The screenshots were featured in a previous slashdot article and they came from Lindows PR.
So we know what their secret to running windows apps is: Wine.
What I find unsettling is that this sneak preview states 99% of Windows applications go "poof" on install (which is not very surprising for anyone who has ever used wine). Yet we've seen screenshots showing Internet Explorer on top of KDE.
So another question stands: are those screenshots real?
In awk put the {print "hello world"} in the BEGIN block. In your case that can be done just by writing BEGIN before the '{'.
:)
Awk executes the BEGIN block unconditionally so it'll print the string regardless of any input (awk's really cool!
With lower salaries, unless you can really prove yourself. I doubt the Ask Slashdot guy will be able to do that since he's so young.
A college degree is better of course, but you're really wrong about needing one.
I wouldn't like to be out of work, 20 years old, with 5 years of questionable industry experience and WITHOUT a college degree.
You're just saying that because of the following:
As you move along in your career a degree matters less and less however. If you are 30 and have proven yourself with a bunch of good years of experience behind you, 95% of 'professionals' will actually respect you, whether or not you got that piece of paper 10 years ago.
You're correct that as you prove yourself the degree becomes less important. Until you've done that, however, you may live through some unpleasant times. If I were 20 years old and in his place, I'd try to go to school. It's much more pleasant than going against the stablished order and gives you better odds.
You have a good point, but I don't believe college is overrated.
I agree with you that college won't turn a "commoner" into a computer wizard and you as an employer are usually looking for the wizard types. The IT market usually looks for technical skills that are usually NOT taught in college. That's because one can learn those on his own. It's much more difficult to learn math on your own than it is with a teacher and that's why colleges may choose to teach math.
College may be overrated in the sense tons of college graduates (the ones who haven't got "it") go out unprepared into the real world. But that doesn't mean we should all ditch college! The unprepared ones can pick up the technical stuff they lack as they work in the real world. They'll still retain the theory they learnt, which will eventually help them in the future. Of course they still won't have "it", but that's no surprise. Most people (CS graduates or not) don't.
We'd both rather have the gifted employee who doesn't need college because he's brilliant. However, I'd rather have the brilliant type who went to college than the brilliant type who didn't. College gives you background which allows you to make better decisions because you're informed. No matter how much you've got "it", you won't be able to figure out the last 100 years of computer science without learning it from books. I believe college is the best place to do your learning.
I agree with you that having "it" is more important than anything else. That's what makes a good CS professional. What I wrote in the original post is advice to the Ask Slashdot guy. Not everybody out there agrees with you and me, and I believe he'll be much more stable with a degree.
[In short: he may be in some trouble if gets fired and he doesn't have a degree. That's why I recommend him to get one while he can. I wish we could all open our little software shops and code away without ever needing to prove ourselves with certificates of learning, but most of us can't.]
It seems to me like this is the "Hi, I'm 19-24, never went to college, am a kick ass programmer with TONS of experience but am stuck professionally! HELP!!!!" week.
I don't intend to offend you, but you must know some things:
1. Face it: college is IMPORTANT. Go study now while you still can. You may argue that you're married with kids now, but you're still young -- don't wait until you're 30 to say "I should've gone to college back when I was 20 and could".
2. Programming is cool, but it's not everything that matters. Once you learn calculus, statistics and logic you'll have better ways to solve problems. You'll see that there are solutions other than brute force.
3. Programming can be extremely easy next to mathematics. It's often much simpler to devise something that "just works" rather than developing it carefully and proving why it's the best solution.
4. Cutting and pasting perl/php for 5 years doesn't count as real work. Nor does assembling computers/networks. Just installing and updating software doesn't count as professional system administration.
5. The computer stuff you did in high school doesn't count as real world experience.
6. Whether you like it or not, most companies will NEVER consider you and most professionals won't respect you if you don't have a degree. You will keep losing arguments even though you're right.
7. You'll never know how much you lack unless you go to college.
In short, you may be stuck professionally for a reason other than age discrimination. Perhaps being a kick ass programmer isn't all that matters.
> As editors, they have a RESPONSIBLITY to get _news_ to us, not their own biased point of view.
No they don't. They can do whatever the hell they want to, because it's their site.
Slashdot is NOT a news agency, and even if it were, they don't have to answer to you, me or anyone else.
Bad news for UNIX is always something minor. That's justified, because the diversity among UNIX machines is higher (meaning bugs rarely affect a lot of people at once) and because UNIX people are more competent.
I'd say the most significant recent UNIX-related exploit was SSH's. (The one which gave free access to accounts with '*' in their password entries). And that's just UNIX-related, because SSH isn't UNIX. Furthermore, the bug doesn't affect OpenSSH, which everyone should be using anyway.
Microsoft bugs, on the other hand, pop up every month and are HUGE, like amazingly virulent worms like Code Red, IE exploits and the recent Windows XP one.
Heh, I can relate.
Once I was at the mall and saw a surveillance camera. Suddenly I had the impulse to pull out a pistol and shoot it.
At that point I realized I'd been playing way too much System Shock II =)
Flavio
I've never used AtheOS and before giving it a try I'd like to see information regarding its hardware requirements and performance. For example, how does AtheOS perform as a web/ftp/samba/db server, considering it supports multithreading and SMP?
Do you have any numbers to show us?
Thanks.
Flavio
Who would you recommend Atheos to?
In other words, where could Atheos be a better choice than other open source OSs?
Thanks!
Flavio
Yeah, I agree. What the hell is the matter with these people?
"Your Rights Online"? Seriously they can't think their rights are being violated. Is there a right to steal?
I don't want to be mean, but I also have a rev. A iMac with 96 MB of RAM. It runs Linux though, and it's blazing fast.
Never swaps and is extremely responsive, even when I run mozilla on it (okay, perhaps not THAT responsive in this case). And I'm also very excited about the possibilities of perl on Linux. (the dev tools also work fine.)
You're yet another one who's distorting reality, omitting facts or making proofless assertions ("the money isn't in software") to make your point logical.
I know some people who have cable connections and CD burners. The ones who know what MP3s are usually download them. Most of these people _never_ buy CDs. The same goes with software.
I'll call it a fact that people who CAN pirate WILL pirate because I have seen this much more often than not.
And the money IS in software. In absolute numbers, much more than in hardware.
And learn how to fucking write, man.
YOU are full of shit because you choose to ignore the status quo to make your little liberty manifesto logical.
Most people who buy pirated games won't buy games legally, even if they're worth retail price.
And considering how much trouble it is to produce, promote and distribute a game I wouldn't say $50's unfair for a good product.
Flavio