Re:Programmers Make Computers Slower Year by Year
on
Netscape 6 Vs. 4.7x
·
· Score: 2
Two comments.
One, the solution to your "problem" can be easily solved by not upgrading. There's noone forcing you to use BeOS 4 over BeOS 3, and you can run Netscape 2.x for all we care. Now it's true that sometimes there are security fixes that aren't backported, but oh well. You have to make a comprimise: Life is full of them.
Two, learning assembly isn't the solution to the problem you describe. In the world of PCs, especially in standard desktop programs, portability and ease of maintence is more important than the bit of performance gained from asm. Hence it makes sense to use a language such as C, C++, or even LISP or Java. Now, I see that you're saying that learning asm is valuable to being able to write good code in any of the languages, as learning how the processor moves bits around is educational. That's true only to a very, very small degree. You have to remember that different processors calculate things differently. Your "Optimized C" may improve perfmorance on x86 by 10%, but watch it tank on a PPC or SPARC. There's a good reason for the abstraction afforded to you by C, and it's not just portability. It's the concept of being able to write code to be readable and portable, without significant regard to the processor it's running on. So while it doesn't hurt to learn the basics, such as it will take a few instruction overhead to change the position of executing (function call) or a few instructions to manipulate some bytes, anything beyond that and you'll have programmers writing worse code.
My experience with asm is limited, mind you. I've done work with Z80 microcontrollers on FOX Trainers (the big black things) and that's roughly it for authoring asm. I've read some x86 asm, and I don't have much interest in it. I've also read some PIC asm, and probably will learn it soon enough. But I do know enough asm to write good C from my experience with a Z80; a processor who's instruction set can't easily be compared to that of x86.
So remember kids: C isn't asm, and don't treat it as such!
For example, you can get closed-source drivers from NVidia, Sigma Designs, Aureal, and dozens of other companies for Linux... but not BSD. You can get Oracle and Wordperfect for Linux, but not BSD.
Drivers are released because hardware manufacturers want people to buy their
hardware. It has nothing to do with the GPL: Windows drivers are free and Windows EULA is a far cry from the GPL.
Oracle and Wordperfect are both written from scratch. The release of them for Linux only is to reduce Q&A by targeting the more popular and buzz-word OS: and also knowing that a Linux binary can run on all modern BSDs.
Frankly, the attention Linux is receiving is not because of it's license. The cases of people contributing back to a project because they are legally bound to are rare; the GPL scares away more companies than it does attract. And those it does attract (read: Linux kernel) do so because it's in their best interest to work on it (hardware manufacturers) or because it's good PR.
So in this case you don't really gain anything from the GPL. Whistle (IBM) and Apple (via Darwin) are the two most prominent examples of companies giving back. And who's to say if they would have been involved if the license was more restrictive?
Re:Why not electronic toys to poor children?
on
Geek Charities?
·
· Score: 1
They've done a good thing, but then they spend the other 355 days of the year pretending the problem isn't there.
Charity is done to make the giver feel good, not to really solve any problems. And that annoys me.
Re:Why not electronic toys to poor children?
on
Geek Charities?
·
· Score: 1
I'm sure the starving children throughout the world prefer a Furby to parents, love, food, or shelter. That's a cop-out gift.
And don't get me started on people who help the poor once (or maybe twice) a year: Thanksgiving, and to a lesser extent, Christmas.
I couldn't agree with you more. While we're at it, let's keep all of the Jews, Blacks, and Hispanics out too. We wouldn't want the issue of genocide brought up, and we wouldn't want the issue of slavery or oppression brought up. These kids are too young for that.
Also, we all know that while there are bad Whites and good Blacks/Hispanics/Jews, we wouldn't want to risk it and let them in. After all, our mantra is Be Prepared; so let's ban everyone who isn't like us. Heck, let's deport them! Be Prepared against Blacks/Hispanics overpowering us!
Surely you aren't suggesting that electronic satelites won't
work; they've been in operation since Sputnik!
Amateur satelites are not news either, with the first being OSCAR 1 launched in 1961!
Yes, I am aware that OSCAR 1 did not last but a few weeks, but there are quite a few QSLs done on it
and OSCAR 2 followed soon after. The reason for the failure of both
were a dead batteries; these satelites did not recharge from solar energy. After all,
they were made by amateurs in garages, etc.
So yes, we do have the technology to send electronic satelites into space: and we've been doing so for nearly 40 years.
Anyhow, it's understandable for stuff like this to slip through.
It's a pretty interesting hack, and certainly one that I'm surpised he got working the first try;) The reason why the atari's motherboard was not smaller to begin with are interesting; generally put you don't use long tracks for your health due to stray capacitance and all... And it wouldn't seem that RF is an issue from surrounding components because well, it's mainly digital.. and it worked for this guy:)
Look at it on the flipside- they already have to take some reasonable effort to keep kids from going to their site.
By classifying their address as.xxx, they may be able to provide services without an age verification system (AVS)... but that does depend on a lot of precedent, yes.
Radio geeks, such as myself, are an old breed that's as geeky, or geekier, than the rest of you.
Sure, shortwave DXing is the most common, but we've all dabbled with medium-wave DXing and FM broadcast band DXing as well.
Perhaps you're saying that not many of us listen to the broadcast bands of the radio for music, and that's only slightly true.. but, don't forget, radio listening encompasses a lot more than listening to your car radio on the way
to work.
He may have been trolling, but he brings up a valid point. The "Screw Echelon" day or whatever it was called was a [pitiful] attempt at a DDoS.
So, CmdrTaco, apparently you do find it justfied sometimes. And so does the attacker of OPN. What's the difference? Oh, you're biased towards Debian, but besides that;)
Well that 4.0 CD probably has source on it, which makes it not terribly difficult to upgrade.
4.0 is fairly recent for a CD starter. My latest CD around here is 3.2, and I use that only when I must.. it's just really easy to upgrade from there, or to install FTP to begin with.
For most purposes, 4.0 is a great introduction. As for only having one book, as another poster pointed out, the concern should be on quality, not quantity.
Oh, and neither Linux or FreeBSD (or any other BSD) are married to 32-bit CPUs;P
yes, but Darwin has a microkernel, but it just runs the BSD kernel as if it was one big subsystem- meaning it's a microkernel architecture, but the kernel that does all of the heavy lifting (everything but what Mach handles) is monolithic.
In other words, if the BSD kernel goes down, you're left with an unusable system... that might as well not be running.
That performance argument is weak. First of all, I don't know if you can inflate Mach's VM enough to call it better than FreeBSD's current VM. In fact, my money rides on FreeBSD's.
Secondly, there's the expense of communication between the microkernel and it's "process", the FreeBSD kernel.
When you've got a microkernel that's only being used to run one monolithic kernel, there goes most of the advantages (to the user, not the programmer) of a microkernel.
Hardly! The Electoral college is one of the few things in our nation that a plurality doesn't totally fib up.
When there is not a majority in the electoral college, the House must decide. Sure it's not perfect, but it's something.
It's the _state_ voting policies, in which they determine which electors they send, that could use the work. The all-or-none systems used in what, 47 states, is the problem.
It's a low power radio transmitter. Surely you wouldn't have posted an article saying "Canakit sells radio transmitter kit", so they throw in MP3 as a buzzword and everyone buys into it.
The technology has nothing to do with MP3. Much as how people say they use their minidiscs to play MP3s when in actuallty their listening to their MP3s reencoded into ATRAC, the term "MP3" has not hing to do at all with what's really happening.
A headline "Cheap Radio Transmitter" would have been more accurate, and it certainly would NOT have been news. Music over radio is certainly not new technology, and after the transistor became popular and cheap, broadcasting inexpensively became a no-news issue too.
No offense to anyone, but I think that slashdot has been suckered into giving this company free PR.
I'm not saying that he's lying, I'm not that cynical. But I am saying that we should take all first party diagnostics of the situation with a grain of salt.
Doesn't want it to be linked to an earlier comprimse. There would be a lot of egg on his face if the problem was brought to his attention earlier and he didn't fix it.
The Chief Security Officer is trying to cover his ass. Take what he says with a grain of salt.
5.1.2.2.1 (Program startup) says that main shall be defined with a return type of int.
It doesn't bring up the issue again until
5.1.2.2.3 (Program termination) where it says if the return type of main is not compatible with int, the behavior is unspecified.
Unspecified means one time it could format your harddrive, another time it could write you a check, and the behavior needs not be documented;)
Actually, it's not illegal, but it's behavior is implementation specific (as opposed to undefined, that is).
I'll open up N869 later and see if I'm wrong, but IIRC it's just not recommended because of the implementation specific behavior- which for most intents and purposes means you shouldn't use it;)
Zork isn't written in C, it's written in Z-Code. Before the invention of the "Z-machine" and consequently Z-Code, Zork's father, dungeon, was written in Fortran.
Activision was never history. They've been that big conglomorate pumping out games forever. Are you confusing them with the now-defunct Infocom, that Activision bought? Activision may still use the name Infocom for some things, but believe me, the real Infocom is long gone, and may it rest in peace.:(
Slashdot Mirror
One, the solution to your "problem" can be easily solved by not upgrading. There's noone forcing you to use BeOS 4 over BeOS 3, and you can run Netscape 2.x for all we care. Now it's true that sometimes there are security fixes that aren't backported, but oh well. You have to make a comprimise: Life is full of them.
Two, learning assembly isn't the solution to the problem you describe. In the world of PCs, especially in standard desktop programs, portability and ease of maintence is more important than the bit of performance gained from asm. Hence it makes sense to use a language such as C, C++, or even LISP or Java. Now, I see that you're saying that learning asm is valuable to being able to write good code in any of the languages, as learning how the processor moves bits around is educational. That's true only to a very, very small degree. You have to remember that different processors calculate things differently. Your "Optimized C" may improve perfmorance on x86 by 10%, but watch it tank on a PPC or SPARC. There's a good reason for the abstraction afforded to you by C, and it's not just portability. It's the concept of being able to write code to be readable and portable, without significant regard to the processor it's running on. So while it doesn't hurt to learn the basics, such as it will take a few instruction overhead to change the position of executing (function call) or a few instructions to manipulate some bytes, anything beyond that and you'll have programmers writing worse code.
My experience with asm is limited, mind you. I've done work with Z80 microcontrollers on FOX Trainers (the big black things) and that's roughly it for authoring asm. I've read some x86 asm, and I don't have much interest in it. I've also read some PIC asm, and probably will learn it soon enough. But I do know enough asm to write good C from my experience with a Z80; a processor who's instruction set can't easily be compared to that of x86.
So remember kids: C isn't asm, and don't treat it as such!
Drivers are released because hardware manufacturers want people to buy their hardware. It has nothing to do with the GPL: Windows drivers are free and Windows EULA is a far cry from the GPL.
Oracle and Wordperfect are both written from scratch. The release of them for Linux only is to reduce Q&A by targeting the more popular and buzz-word OS: and also knowing that a Linux binary can run on all modern BSDs.
Frankly, the attention Linux is receiving is not because of it's license. The cases of people contributing back to a project because they are legally bound to are rare; the GPL scares away more companies than it does attract. And those it does attract (read: Linux kernel) do so because it's in their best interest to work on it (hardware manufacturers) or because it's good PR.
So in this case you don't really gain anything from the GPL. Whistle (IBM) and Apple (via Darwin) are the two most prominent examples of companies giving back. And who's to say if they would have been involved if the license was more restrictive?
Charity is done to make the giver feel good, not to really solve any problems. And that annoys me.
And don't get me started on people who help the poor once (or maybe twice) a year: Thanksgiving, and to a lesser extent, Christmas.
Also, we all know that while there are bad Whites and good Blacks/Hispanics/Jews, we wouldn't want to risk it and let them in. After all, our mantra is Be Prepared; so let's ban everyone who isn't like us. Heck, let's deport them! Be Prepared against Blacks/Hispanics overpowering us!
With $17,000 on the line, I don't think this guy is about to commit mail fraud. To be frank, he wouldn't get away with it.
That said, a unified ABI may not be too far away. Who knows?
Amateur satelites are not news either, with the first being OSCAR 1 launched in 1961!
Yes, I am aware that OSCAR 1 did not last but a few weeks, but there are quite a few QSLs done on it and OSCAR 2 followed soon after. The reason for the failure of both were a dead batteries; these satelites did not recharge from solar energy. After all,
So yes, we do have the technology to send electronic satelites into space: and we've been doing so for nearly 40 years.
It's a pretty interesting hack, and certainly one that I'm surpised he got working the first try ;) The reason why the atari's motherboard was not smaller to begin with are interesting; generally put you don't use long tracks for your health due to stray capacitance and all... And it wouldn't seem that RF is an issue from surrounding components because well, it's mainly digital.. and it worked for this guy :)
By classifying their address as .xxx, they may be able to provide services without an age verification system (AVS)... but that does depend on a lot of precedent, yes.
Poll: What crashes most?
Windows 9%
Netscape 10%
Hemos 70%
(Yes, 11% of the vote is mysteriously unaccounted for)
Radio geeks, such as myself, are an old breed that's as geeky, or geekier, than the rest of you.
Sure, shortwave DXing is the most common, but we've all dabbled with medium-wave DXing and FM broadcast band DXing as well.
Perhaps you're saying that not many of us listen to the broadcast bands of the radio for music, and that's only slightly true.. but, don't forget, radio listening encompasses a lot more than listening to your car radio on the way to work.
So, CmdrTaco, apparently you do find it justfied sometimes. And so does the attacker of OPN. What's the difference? Oh, you're biased towards Debian, but besides that ;)
Grown men have gone to war on issues less important. Don't be quick to blame the attack on a child.
4.0 is fairly recent for a CD starter. My latest CD around here is 3.2, and I use that only when I must.. it's just really easy to upgrade from there, or to install FTP to begin with.
For most purposes, 4.0 is a great introduction. As for only having one book, as another poster pointed out, the concern should be on quality, not quantity.
Oh, and neither Linux or FreeBSD (or any other BSD) are married to 32-bit CPUs ;P
In other words, if the BSD kernel goes down, you're left with an unusable system... that might as well not be running.
Secondly, there's the expense of communication between the microkernel and it's "process", the FreeBSD kernel.
When you've got a microkernel that's only being used to run one monolithic kernel, there goes most of the advantages (to the user, not the programmer) of a microkernel.
When there is not a majority in the electoral college, the House must decide. Sure it's not perfect, but it's something.
It's the _state_ voting policies, in which they determine which electors they send, that could use the work. The all-or-none systems used in what, 47 states, is the problem.
www.slashdto.org
www.slashdt.org
www.sashdot.org
www.slshdot.org
www.slsahdot.org
www.slasdot.org
www.salhdot.org
There's perhaps a few more, I'm tired and don't feel like wasting more time investigating ;)
The technology has nothing to do with MP3. Much as how people say they use their minidiscs to play MP3s when in actuallty their listening to their MP3s reencoded into ATRAC, the term "MP3" has not hing to do at all with what's really happening.
A headline "Cheap Radio Transmitter" would have been more accurate, and it certainly would NOT have been news. Music over radio is certainly not new technology, and after the transistor became popular and cheap, broadcasting inexpensively became a no-news issue too.
No offense to anyone, but I think that slashdot has been suckered into giving this company free PR.
I'm not saying that he's lying, I'm not that cynical. But I am saying that we should take all first party diagnostics of the situation with a grain of salt.
The Chief Security Officer is trying to cover his ass. Take what he says with a grain of salt.
5.1.2.2.1 (Program startup) says that main shall be defined with a return type of int.
It doesn't bring up the issue again until 5.1.2.2.3 (Program termination) where it says if the return type of main is not compatible with int, the behavior is unspecified.
Unspecified means one time it could format your harddrive, another time it could write you a check, and the behavior needs not be documented ;)
I'll open up N869 later and see if I'm wrong, but IIRC it's just not recommended because of the implementation specific behavior- which for most intents and purposes means you shouldn't use it ;)
Activision was never history. They've been that big conglomorate pumping out games forever. Are you confusing them with the now-defunct Infocom, that Activision bought? Activision may still use the name Infocom for some things, but believe me, the real Infocom is long gone, and may it rest in peace. :(