I don't think this is necessarly flaimbait...
The FSF does have a lot of zealots. That's why a lot of people don't GPL their code. There are eleventy twelve bazillion licenses that are similar to the GPL. And even more that aren't. FSF is very politically motivated and is interested in a huge movement. Some people just want to write software and don't want to get involved in all of that.
A bit off topic but....
I was under the impression the BSD license was GPL compatable. I loaded a BSD licensed kernel module today (Intel iscsi) and up popped a message saying my kernel was tainted and to not expect support. I could understand no support... but I thought BSD license was compatable. Why did it then taint my kernel?
I think the thread is more about Intel's attempt's to call generialized technology an Intel marketing name. ia32 instead of x86 etc. etc.
Personally, I'd rather see more efficient chips like that of SPARC's and G5's. 64-bit really doesn't benefit that many people at all. Instead of bigger and faster and more mhz, why not smaller and cooler? The problem isn't chip clock cycles or how much memory space a program can see at once... the problem is latency and retardedly long pipelines. Oh well, the Intel marketing machine at work!
Yeah, that's bullshit. I know people who run small isps and hosting environments, and if they suffered anything like that, the cost of downtime would probably put their business in serious trouble. Oh, and who do the customers complain to? Not the FBI, that's for sure! Unless someone's life was in danger, there's no need to hurt someone's business like that. Yeah, fuck a bunch of FBI. Let's just hope someone pees in their coffee.
I know employers whom have posted jobs on monster, and they get such a crap flood of responses your resume is almost guarnteed to get lost in the mess. Even if it is good, the number of applicants they get is just overwhelming. I know of a local position that posted on monster and got 300 applicants in a short period of time. This dwarfs all the other applications they had gotten before that.
My favorite way to get a job is still good old fashioned networking. Many people in the IT industry are... non-sociable to say the least (by that I don't mean they hide in their rooms all day, but a lot of them are elitists etc. etc.). So if your out there meeting and greeting you've got a huge competitive edge.
I tend to agree. Solaris isn't a bad operating system... the need for it is fading though. Who _needs_ Solaris boxes on Sun hardware anymore? x86 servers can reasonably serve thousands and thousands of users these days. Sun hardware just isn't as attractive as it once was. I think as far as database servers they still have a very nice competitive edge, but as x86 hardware becomes more powerful and gets more of the features of Sun hardware (lower latency) we will see Sun hardware become even less pervasive. Sun boxes are nice, but most really can't justify the cost.
As gay as the name is, I think this is their way of protest. There are a million and one product names that sound similar to windows. Winamp, winfax, windildo. I think it's total 100% bullshit you can selectivly prosecute only those who have competing products. There's a computer repair company in my town called wintech. How MS can selectively sue trademark infringements based on their level of competition and say they aren't a monopoly with a straight face is beyond me.
Not that I really think MS would ever touch linux with a ten foot pole, but....
a) MS always keeps things hush hush. So their denial doesn't really mean that much.
b) It makes sense they would want to snatch up the office arena of linux before open office gets any better. If they can have people using MS Office on Linux, the likely hood of people using open office becomes lower. This is very bad. There will still be a certain crowd still using it because it's open source, but there will be a crowd who switches over. This means less people using it, reporting bugs, trying to make it better, etc. etc.
I really wouldn't necessarly rule out Office on linux simply for the fact there's office on the Mac. The difference though is Mac's and Windows have two different core audiences. On the other hand, Windows and Linux do have overlapping users far greater than that of Mac/Windows.
Oh, and btw, a bit OT, but... why isn't there high level scripting language support (perl, python, ruby, etc. etc.) for open office? (or if there is someone please point me toward it!). OO would be the perfect canidate for it. Mozilla has XUL, gaim has perl/c. It would definatly put OO ahead of the pack.
Microsoft has a free publicly released ram disk driver. It's a complete piece of crap. It has like an 8MB limit. That's far too small to do anything with. And to resize it you have to reboot. There are some other commercial 3rd party ram disk drivers out there, but they are expensive (usually over 100 dollars, sometimes over 500). Windows should have an included ram disk driver. It's an EXTREMELY useful tool for both servers and desktops.
>Actually, the actual Nt_ interfaces *are* documented, but (afaik) incompletely
You are correct sir. Yes!
Really though, the basic rule of thumb is, everything MSDN doc related is well documented for basic things, but advanced things are imcompletely documented, documents are out of date, or not documentated at all. A perfect example was of a defragmenting program for NTFS. NTFS at the time (or maybe even still today) was not fully documented, so this defrag product was corupting files on people's hard drives. It was later found out that an undocumented feature/bug was the culprit (was a 3rd party defrag program).
I know a few people who've ripped their hair out for a few hours trying to figure out why certain things weren't "acting" right only to find out it was something incompletely documented.
Many popular games that were written more around opengl than directx are ported. I was just playing UT2004 demo in linux this afternoon. I don't think there's a huge demand for linux games anyway. On the list of things to do, that's like number 4,563. "Never linux" is a far far cry, and I'd say the numbers are pretty close to equal these days. If a company is going to take the time to port, why not double your ported OS audience.
Also, does anyone know how these studies are done? Please tell me they are not basing it on store bought versions of linux. Every single person I know personally using linux has downloaded an iso and installed it that way. Even the people with limited bandwidth will just borrow someone else's burned CD.
There is actually a lot of network related code in there. Microsoft while trying to downplay, it can't deny that 13 million lines have been released. It doesn't matter the total size of windows and whether this is 1% or 25%. The old addage is you can count on one mistake for every thousand lines of code. Look at programs that are just a few thousand lines of code that have exploits. I'd say at the very least, we are looking at 20 buffer overflows in the code. Obviously not every single one will be found, but you can count on a few. Espically since people will be looking mighty hard. With comments like "this may be off by -1, but I'm not sure", I think we are almost guarenteed some buffer overflows.
This will also give the daring souls willing to look at it a chance to tell us if there is GPL code. Rumor is GNU style Makfiles (which isn't illegal) and parts of gnu autoconf (which I suspect is illegal, if they actually include it in the OS).
Hmmmm....
For perl code to be included Windows, it would mean they would have to have a perl interpreter in windows (It's highly doubtful they used perlcc). Does the perl license similar to the BSD license where they could legally put the offical perl in windows, or is it more like a GNU style license where they couldn't.
At this point it's way too late. There are foreign websites hosting it, and it's going to be next to impossible for MS to have any pull on some of these countries.
I remember a high ranking Microsoft offical once stated that if the windows source code ever got out, it is so bug filled it would be devastating. The part I find funny about this is no one even questions that holes will be found and worms will be written. It's just a question of how many.
The playing field is even though when it comes to backdoors. Yeah, debian and FSF and OpenSSH have been cracked and backdoors attempted at one point or another. but, the same thing happens to closed source software. It's just different. Half Life 2 could have been back doored if the cracker wanted to go that route. There have been many backdoors that have snuck into programs by employees. You can argue all day as to why one is better than the other, but it comes down to, they are just different.
One of the best things we have to fight against this is gpg sigs and md5 sums. I know they aren't perfect because if someone cracked the server those are kept on then it doesn't really matter. but none the less, they do help a lot; and distros that have made it a point to check them before installing have helped lessen the problem of backdoored software.
A funny observation... it seems on slashdot any time you put a conflicting view point up (conflicting to the slashdot norm) it gets modded down, but the second you put "I am not trolling" or "I'll probably get modded down for this.." it gets modded up. Just an interesting observation.
"Feels faster" as in when I type in text I don't get any delay. Feels faster as when I open an application it opens immediatly rather than a 3-4 second delay. Feels faster as in I can drag applications across the desktop without ghosting effects. Feels faster as in compiling GCC takes 15 minutes rather than 45. Feels faster as in video isn't choppy. Feels faster as in huge mysql queries take a third the time.
"Feels faster" does have basis. I have a feeling you probably haven't tried an i686 distro extensivly or else you probably wouldn't have the same view point.
But it still is considerbly slower. The proof is in the pudding. Compare a i386 distribution to a source one with sse, mmx, -march=athlon-xp etc. etc. It feels considerbly faster (which is more important to me than benchmarks). Not a Gentoo zealot =P but I've tried a few i686 distro's (first one was beehive for those of you that remember) and a bunch source compiled distro's with good compile time options and it's night and day.
I run Linux from Scratch and compile almost everything myself (mozilla binaries work well so I use those). The problem with a standard database, is it's still a database. autoconf doesn't use a database. It actually will check to see if gcc -lm is ok, then see if gzip is there, etc. etc. My "native format" is source. Using a./configure type of environment checking is MUCH more portable and allows for me to install just a few binaries easily, rather than having to build the database manually to install those few packages. Autopackage is actually doing this, and I think autopackage should become the new standard packaging system.
GPG sigs and MD5sums aren't really leaps and bounds. RPM still is basically a cpio wrapper. It still relies on a database with all the packages. If I compiled Xfree86 from source and XFree86 is a dep. for a package, it will not reconize my xfree86 because it's not in the database. We need to get rid of the database.
We still have the problem of, unless it's statically compiled or you symlink the libs yourself to the version it's expecting, it's difficult to have all the exact lib versions the package was compiled against.
Have you ever tried to put RPM on a non-rpm distro? First of all, it's a pain in the ass to compile. From what I remember I had to write a patch for it. Second, you have to build the database of all the packages that you've compiled yourself, manually. Then, for every package you compile yourself, you have to add it to the database. By the time you are done with all that, you could have compiled it from source.
Slashdot Mirror
I don't think this is necessarly flaimbait... The FSF does have a lot of zealots. That's why a lot of people don't GPL their code. There are eleventy twelve bazillion licenses that are similar to the GPL. And even more that aren't. FSF is very politically motivated and is interested in a huge movement. Some people just want to write software and don't want to get involved in all of that.
A bit off topic but....
I was under the impression the BSD license was GPL compatable. I loaded a BSD licensed kernel module today (Intel iscsi) and up popped a message saying my kernel was tainted and to not expect support. I could understand no support... but I thought BSD license was compatable. Why did it then taint my kernel?
I think the thread is more about Intel's attempt's to call generialized technology an Intel marketing name. ia32 instead of x86 etc. etc.
Personally, I'd rather see more efficient chips like that of SPARC's and G5's. 64-bit really doesn't benefit that many people at all. Instead of bigger and faster and more mhz, why not smaller and cooler? The problem isn't chip clock cycles or how much memory space a program can see at once... the problem is latency and retardedly long pipelines. Oh well, the Intel marketing machine at work!
I'd say the submitter was trying to sneak one by. That article was too short and too straight forward to be misinterpreted that badly.
Yeah, that's bullshit. I know people who run small isps and hosting environments, and if they suffered anything like that, the cost of downtime would probably put their business in serious trouble. Oh, and who do the customers complain to? Not the FBI, that's for sure! Unless someone's life was in danger, there's no need to hurt someone's business like that. Yeah, fuck a bunch of FBI. Let's just hope someone pees in their coffee.
I know employers whom have posted jobs on monster, and they get such a crap flood of responses your resume is almost guarnteed to get lost in the mess. Even if it is good, the number of applicants they get is just overwhelming. I know of a local position that posted on monster and got 300 applicants in a short period of time. This dwarfs all the other applications they had gotten before that.
My favorite way to get a job is still good old fashioned networking. Many people in the IT industry are... non-sociable to say the least (by that I don't mean they hide in their rooms all day, but a lot of them are elitists etc. etc.). So if your out there meeting and greeting you've got a huge competitive edge.
I tend to agree. Solaris isn't a bad operating system... the need for it is fading though. Who _needs_ Solaris boxes on Sun hardware anymore? x86 servers can reasonably serve thousands and thousands of users these days. Sun hardware just isn't as attractive as it once was. I think as far as database servers they still have a very nice competitive edge, but as x86 hardware becomes more powerful and gets more of the features of Sun hardware (lower latency) we will see Sun hardware become even less pervasive. Sun boxes are nice, but most really can't justify the cost.
More likely it would be used to erase someone's childhood to store hundreds of GB of sex memories.
It's a self-fufilling prophecy. When google gets it in it's database from slashdot, it will then exist :)
As gay as the name is, I think this is their way of protest. There are a million and one product names that sound similar to windows. Winamp, winfax, windildo. I think it's total 100% bullshit you can selectivly prosecute only those who have competing products. There's a computer repair company in my town called wintech. How MS can selectively sue trademark infringements based on their level of competition and say they aren't a monopoly with a straight face is beyond me.
Not that I really think MS would ever touch linux with a ten foot pole, but....
a) MS always keeps things hush hush. So their denial doesn't really mean that much.
b) It makes sense they would want to snatch up the office arena of linux before open office gets any better. If they can have people using MS Office on Linux, the likely hood of people using open office becomes lower. This is very bad. There will still be a certain crowd still using it because it's open source, but there will be a crowd who switches over. This means less people using it, reporting bugs, trying to make it better, etc. etc.
I really wouldn't necessarly rule out Office on linux simply for the fact there's office on the Mac. The difference though is Mac's and Windows have two different core audiences. On the other hand, Windows and Linux do have overlapping users far greater than that of Mac/Windows.
Oh, and btw, a bit OT, but... why isn't there high level scripting language support (perl, python, ruby, etc. etc.) for open office? (or if there is someone please point me toward it!). OO would be the perfect canidate for it. Mozilla has XUL, gaim has perl/c. It would definatly put OO ahead of the pack.
Microsoft has a free publicly released ram disk driver. It's a complete piece of crap. It has like an 8MB limit. That's far too small to do anything with. And to resize it you have to reboot. There are some other commercial 3rd party ram disk drivers out there, but they are expensive (usually over 100 dollars, sometimes over 500). Windows should have an included ram disk driver. It's an EXTREMELY useful tool for both servers and desktops.
Get some damn common.
>Actually, the actual Nt_ interfaces *are* documented, but (afaik) incompletely
You are correct sir. Yes! Really though, the basic rule of thumb is, everything MSDN doc related is well documented for basic things, but advanced things are imcompletely documented, documents are out of date, or not documentated at all. A perfect example was of a defragmenting program for NTFS. NTFS at the time (or maybe even still today) was not fully documented, so this defrag product was corupting files on people's hard drives. It was later found out that an undocumented feature/bug was the culprit (was a 3rd party defrag program).
I know a few people who've ripped their hair out for a few hours trying to figure out why certain things weren't "acting" right only to find out it was something incompletely documented.
Many popular games that were written more around opengl than directx are ported. I was just playing UT2004 demo in linux this afternoon. I don't think there's a huge demand for linux games anyway. On the list of things to do, that's like number 4,563. "Never linux" is a far far cry, and I'd say the numbers are pretty close to equal these days. If a company is going to take the time to port, why not double your ported OS audience.
Also, does anyone know how these studies are done? Please tell me they are not basing it on store bought versions of linux. Every single person I know personally using linux has downloaded an iso and installed it that way. Even the people with limited bandwidth will just borrow someone else's burned CD.
There's no NTFS code in there.
There is actually a lot of network related code in there. Microsoft while trying to downplay, it can't deny that 13 million lines have been released. It doesn't matter the total size of windows and whether this is 1% or 25%. The old addage is you can count on one mistake for every thousand lines of code. Look at programs that are just a few thousand lines of code that have exploits. I'd say at the very least, we are looking at 20 buffer overflows in the code. Obviously not every single one will be found, but you can count on a few. Espically since people will be looking mighty hard. With comments like "this may be off by -1, but I'm not sure", I think we are almost guarenteed some buffer overflows.
This will also give the daring souls willing to look at it a chance to tell us if there is GPL code. Rumor is GNU style Makfiles (which isn't illegal) and parts of gnu autoconf (which I suspect is illegal, if they actually include it in the OS).
Oh the irony! Windows source code becomes leaked because the operating system is was running on (windows) wasn't secure.
Hmmmm....
For perl code to be included Windows, it would mean they would have to have a perl interpreter in windows (It's highly doubtful they used perlcc). Does the perl license similar to the BSD license where they could legally put the offical perl in windows, or is it more like a GNU style license where they couldn't.
At this point it's way too late. There are foreign websites hosting it, and it's going to be next to impossible for MS to have any pull on some of these countries.
A bit hypocritical for them to use autoconf and GNU style makefiles don't ya think?
I remember a high ranking Microsoft offical once stated that if the windows source code ever got out, it is so bug filled it would be devastating. The part I find funny about this is no one even questions that holes will be found and worms will be written. It's just a question of how many.
The playing field is even though when it comes to backdoors. Yeah, debian and FSF and OpenSSH have been cracked and backdoors attempted at one point or another. but, the same thing happens to closed source software. It's just different. Half Life 2 could have been back doored if the cracker wanted to go that route. There have been many backdoors that have snuck into programs by employees. You can argue all day as to why one is better than the other, but it comes down to, they are just different.
One of the best things we have to fight against this is gpg sigs and md5 sums. I know they aren't perfect because if someone cracked the server those are kept on then it doesn't really matter. but none the less, they do help a lot; and distros that have made it a point to check them before installing have helped lessen the problem of backdoored software.
A funny observation... it seems on slashdot any time you put a conflicting view point up (conflicting to the slashdot norm) it gets modded down, but the second you put "I am not trolling" or "I'll probably get modded down for this.." it gets modded up. Just an interesting observation.
"Feels faster" as in when I type in text I don't get any delay. Feels faster as when I open an application it opens immediatly rather than a 3-4 second delay. Feels faster as in I can drag applications across the desktop without ghosting effects. Feels faster as in compiling GCC takes 15 minutes rather than 45. Feels faster as in video isn't choppy. Feels faster as in huge mysql queries take a third the time.
"Feels faster" does have basis. I have a feeling you probably haven't tried an i686 distro extensivly or else you probably wouldn't have the same view point.
But it still is considerbly slower. The proof is in the pudding. Compare a i386 distribution to a source one with sse, mmx, -march=athlon-xp etc. etc. It feels considerbly faster (which is more important to me than benchmarks). Not a Gentoo zealot =P but I've tried a few i686 distro's (first one was beehive for those of you that remember) and a bunch source compiled distro's with good compile time options and it's night and day.
I run Linux from Scratch and compile almost everything myself (mozilla binaries work well so I use those). The problem with a standard database, is it's still a database. autoconf doesn't use a database. It actually will check to see if gcc -lm is ok, then see if gzip is there, etc. etc. My "native format" is source. Using a ./configure type of environment checking is MUCH more portable and allows for me to install just a few binaries easily, rather than having to build the database manually to install those few packages. Autopackage is actually doing this, and I think autopackage should become the new standard packaging system.
GPG sigs and MD5sums aren't really leaps and bounds. RPM still is basically a cpio wrapper. It still relies on a database with all the packages. If I compiled Xfree86 from source and XFree86 is a dep. for a package, it will not reconize my xfree86 because it's not in the database. We need to get rid of the database.
We still have the problem of, unless it's statically compiled or you symlink the libs yourself to the version it's expecting, it's difficult to have all the exact lib versions the package was compiled against.
Have you ever tried to put RPM on a non-rpm distro? First of all, it's a pain in the ass to compile. From what I remember I had to write a patch for it. Second, you have to build the database of all the packages that you've compiled yourself, manually. Then, for every package you compile yourself, you have to add it to the database. By the time you are done with all that, you could have compiled it from source.