You don't need to look at the code to write an ext2fs clone.
And for that matter, you could look at the Linux code and then write your own Ext2fs implementation. There's no such thing as 'contamination' just from reading GPL'd Linux source code.
If the GPL is completely invalid, more than SCO have a 'singular problem.'
It becomes impossible for ANYBODY to distribute GPL'd code. The Linux kernal source tree would magically be transformed into a big 'gimme gimme' festival as every contributor tried to figure out what the new standing was of his/her contributed components therein.
What really really clogs up the discussions on this website are tards like you who use your +2 default to spew offtopic crap about 'moderation issues', and then are even lame enough to spam us with corrections to the typos in your offtopic blather with your +2 default.
There really needs to be a way for 'the moderation system' to slap crap like that down hard, i.e. make you lose your +2 posting privledges for not having a CLUE that moderation is about improving comment quality, NOT a little contest to boost ego and give out cookies to the best commenters.
Yikes. There's my meta-meta off-topic rant-about-moderation.
Heck, J5 didn't even have an entire single gigabyte of hard drive space, and my pocket calculator has more RAM than J5 did.
Yes, but your pocket calculator is probably programmed in C++ or some other hideously inefficient programming language. A gigabyte of storage is actually one hell of a lot of space. 64M of RAM is more than enough for a hell of a lot of optimized assembly code and compressed data.
Remember, the huge amounts of RAM and storage in hardware these days is to follow the Microsoft Model of software development: Roll in crap to fill all available storage. Some of us still (believe it or not) feel that software development is convergent. That means: as software gets better and more sophisticated, it should be implementing more functionality while consuming less resources. i.e.: the 486 with 16 megs of RAM that I used to run Slackware on should be BETTER at running the latest Slackware, not worse.
But I'm an optomist. For some reason eye candy is easer to shove into a system than actual increased functionality. So we have shit like GNOME now, not new-improved FVWM3 or FVWM4.
Actually, Apache stopped Netscape from monopolizing the web, too. Netscape had great plans for proprietary HTML extensions and a firm handshake between Netscape clients and servers.
That's one good thing that came out of Microsoft winning the 'web browser' competition against Netscape. And since the end result was that the Netscape 'beast' was tamed and out of the ashes of it came Mozilla, it turned out about as good as it ever could have.
Naw. There's a really good free email client out there. It's ad-supported, but Eudora is a nice email client. Or you can pay a one time registration and never see the ads again.
I'm sure there are plenty of stories to tell. However, most Windows exploits are rather weak. Very few give the cracker a bash prompt, with C Compiler and all the tools that a root shell exploit on a Linux box gives.
There are mirrors of ftp.gnu.org all over the Internet.
The 'damage' done here is much more severe than somebody just deleting everything at the actual ftp.gnu.org site. Because if that had happened, it would have been a simple matter to stream it all back from one of the mirror sites.
What happened instead is far worse. Someone got in and (may have) corrupted some of the source tarballs. Then, because it wasn't noticed immediately, these corrupted files made their way out to the mirrors.
No. A site running Linux that consists of the primary repository for much of the software that makes up Linux was hacked. And there was no backup of the files on the site. And the maintainers of the site have had to resort to a public 'please help us get this all put back together' plea.
And it isn't a matter of 'Let's everyone use Windows.'
Yes, but those are IDE CDR drives, and everybody knows that (for god knows what reason) the maintainer of cdrecord only directly supports SCSI CDR drives. Is it still necessary to use the 'SCSI emulation kludge' in a recompiled kernal to get your IDE CDR to work in Linux? It's been a few years since I tried.
Slackware 3.6, and any version of Slackware before version 4.0 came out, didn't assign a root password, or prompt the installer to assign a root password.
I happened upon a friend's system one night (after determining her IP address by reading a recent mail header) and on a lark telnetted in and typed 'root' as a user name. Bip! I had root access to her Slackware box. She'd been online through a PPP connection for several weeks with her box in that state.
I am sure this is not the only anecdotal evidence of security problems with Linux distributions. There are so many versions of Linux out there now, that I'm sure there are stories to tell with any of them.
With some 'distributions', i.e. Free/Net/OpenBSD, the source tarball for packages is downloaded directly from the source repository at GNU. An MD5sum test is made, of course, but it is NOT a matter of trusting a commercial distributor.
Why do people assume there needs to be a commercial distributor?
I can remember back when I first installed Yggdrasil Linux on my 486 box in 1993 that the emacs install was like 80 megs. There was no way I could afford that much HD space on my tiny 330 meg hard drive.
BSD ports don't consist of source file sets that have been fucked with by the distribution assembler, like Red Hat SRPMs and what-not. The patches are instead applied during the build process to pristine tarballs of the source code from wherever the source originated.
The only XP computers 'suspectable to MSBlast' are those that are unpatched, which is significantly fewer than 'every single machine'. Just the same as the only Linux servers (known) to be susceptible are those without current patches applied. And there are plenty of Linux boxes out there that aren't kept up to date. Nobody told the people who put them online that the 'free' operating system was going to require 10 hours a week of reading Usenet security groups and update websites and applying patches.
Historically, Richard Stallman was one of the hackers at MIT who actively opposed the impostion of passwords on the Unix account logins. He and other hackers like him at the time opposed passwords because they believed in a community of sharing and openness. They refused to put passwords on their accounts for as long as possible.
Yep. And the same exemption-from-liability is common in the Free Software world. I think that right in the COPYING file there's an expressed 'no-liability' clause.
We'd better get right on that. I can see all sorts of free software developers needing liability insurance policies before they can push the new source tarball out on the ftp site.
Apple managed to do it by standing on the shoulders of giants,
Apple managed to do it by pissing away millions and millions of dollars trying to prove they were capable of producing a real OS with real multitasking, robustness, etc. All through the 90's they used up cute code-names on 'next-generation Operating System' buzzword sinkholes (object oriented this-n-that of the year).
Then they finally gave up and just put new greasepaint on NextStep instead. Yep. The NIHNI (not invented here? not interested!) brigade finally just went out and bought an OS because they couldn't roll one of their own.
Boy that had to hurt if you're a middle-aged old-line OS developer at Apple.
Maybe the firewalling should be done at the ISP. Block all incoming traffic to un-needed ports. If someone needs Port 80, open it up and bill accordingly.
Oh, wait. I hear all the screaming and fury coming from a mile away at that idea.
I've yet to find a good Architectural and/or Land Development CADD program for Mac or Linux. Nor Noise simulation modules, Motorola propegation simulators, Hydrology simulations, or many more of the specialized software we use for buidling design, airport/runway design, emergancy system management, wireless design, air quality analysis, or any of the other stuff we do at my company.
There isn't a heck of a lot of good engineering design software for Linux. There never will be in the form of Open Source. It's software that costs $2-30,000 per seat. You know, software for grownups, not dilletantes who browse the web and 'admin' common commodity tasks like web servers. We can't all just sell stuff and/or present it for sale. Somebody has to design it.
Microsoft is already on it. The whole.NET initiative. You didn't think they were going to keep on keeping on with the old ways, did you? Subscription software all the way. They'll make more money that way, and with them in control of your system, the problems seen here will go away, i.e. every one of the problem machines out there right now would already have been patched automatically.
This whole crisis is an opportunity for Microsoft to hype what they've been hyping recently (subscription software, downloaded from MS regularly).
It must have something to do with you having a hell of a lot of bandwidth available to download all that stuff over and over and over again with each x.x.x.01 revision. That won't scale very well to the whole world.
That day will never come. Enough of us are of an age to remember the days when there were fifteen different PC platforms out there and the huge splintered market for commercial software that resulted.
It's trouble enough for retailers to sell both Mac and PC games. Do you really think shrinkwrapped boxes are going to contain the seven CDs necessary to have the app run on 15 seperate OSes?
Yeah, everything will be distributed as source code. Uh-huh. People will like that.
Yep. And The Committee will decide what software you run on your PC. Better go get a Form # 34958/12 and fill it out. In a week or so (maybe) somebody from IT will roll out that executable and you'll be able to do your work.
Fuck that. It might work in a stodgy bureaucratic business enviroment, i.e. in big dinosaur corporate envionments. It will NEVER work in small dynamic companies, let ALONE in a home computing environment.
Slashdot Mirror
You don't need to look at the code to write an ext2fs clone.
And for that matter, you could look at the Linux code and then write your own Ext2fs implementation. There's no such thing as 'contamination' just from reading GPL'd Linux source code.
If the GPL is completely invalid, more than SCO have a 'singular problem.'
It becomes impossible for ANYBODY to distribute GPL'd code. The Linux kernal source tree would magically be transformed into a big 'gimme gimme' festival as every contributor tried to figure out what the new standing was of his/her contributed components therein.
You meant to say NetBSD. Linux can be ported badly to any architecture, with a partial non-standard userland thrown together by someone at random.
With NetBSD if it's ported to an architecture, it's the same identical codebase as on every other NetBSD architecture.
What really really clogs up the discussions on this website are tards like you who use your +2 default to spew offtopic crap about 'moderation issues', and then are even lame enough to spam us with corrections to the typos in your offtopic blather with your +2 default.
There really needs to be a way for 'the moderation system' to slap crap like that down hard, i.e. make you lose your +2 posting privledges for not having a CLUE that moderation is about improving comment quality, NOT a little contest to boost ego and give out cookies to the best commenters.
Yikes. There's my meta-meta off-topic rant-about-moderation.
Heck, J5 didn't even have an entire single gigabyte of hard drive space, and my pocket calculator has more RAM than J5 did.
Yes, but your pocket calculator is probably programmed in C++ or some other hideously inefficient programming language. A gigabyte of storage is actually one hell of a lot of space. 64M of RAM is more than enough for a hell of a lot of optimized assembly code and compressed data.
Remember, the huge amounts of RAM and storage in hardware these days is to follow the Microsoft Model of software development: Roll in crap to fill all available storage. Some of us still (believe it or not) feel that software development is convergent. That means: as software gets better and more sophisticated, it should be implementing more functionality while consuming less resources. i.e.: the 486 with 16 megs of RAM that I used to run Slackware on should be BETTER at running the latest Slackware, not worse.
But I'm an optomist. For some reason eye candy is easer to shove into a system than actual increased functionality. So we have shit like GNOME now, not new-improved FVWM3 or FVWM4.
Actually, Apache stopped Netscape from monopolizing the web, too. Netscape had great plans for proprietary HTML extensions and a firm handshake between Netscape clients and servers.
That's one good thing that came out of Microsoft winning the 'web browser' competition against Netscape. And since the end result was that the Netscape 'beast' was tamed and out of the ashes of it came Mozilla, it turned out about as good as it ever could have.
Naw. There's a really good free email client out there. It's ad-supported, but Eudora is a nice email client. Or you can pay a one time registration and never see the ads again.
I'm sure there are plenty of stories to tell. However, most Windows exploits are rather weak. Very few give the cracker a bash prompt, with C Compiler and all the tools that a root shell exploit on a Linux box gives.
There are mirrors of ftp.gnu.org all over the Internet.
The 'damage' done here is much more severe than somebody just deleting everything at the actual ftp.gnu.org site. Because if that had happened, it would have been a simple matter to stream it all back from one of the mirror sites.
What happened instead is far worse. Someone got in and (may have) corrupted some of the source tarballs. Then, because it wasn't noticed immediately, these corrupted files made their way out to the mirrors.
No. A site running Linux that consists of the primary repository for much of the software that makes up Linux was hacked. And there was no backup of the files on the site. And the maintainers of the site have had to resort to a public 'please help us get this all put back together' plea.
And it isn't a matter of 'Let's everyone use Windows.'
Oh! I think I've been trolled.
Yes, but those are IDE CDR drives, and everybody knows that (for god knows what reason) the maintainer of cdrecord only directly supports SCSI CDR drives. Is it still necessary to use the 'SCSI emulation kludge' in a recompiled kernal to get your IDE CDR to work in Linux? It's been a few years since I tried.
Slackware 3.6, and any version of Slackware before version 4.0 came out, didn't assign a root password, or prompt the installer to assign a root password.
I happened upon a friend's system one night (after determining her IP address by reading a recent mail header) and on a lark telnetted in and typed 'root' as a user name. Bip! I had root access to her Slackware box. She'd been online through a PPP connection for several weeks with her box in that state.
I am sure this is not the only anecdotal evidence of security problems with Linux distributions. There are so many versions of Linux out there now, that I'm sure there are stories to tell with any of them.
With some 'distributions', i.e. Free/Net/OpenBSD, the source tarball for packages is downloaded directly from the source repository at GNU. An MD5sum test is made, of course, but it is NOT a matter of trusting a commercial distributor.
Why do people assume there needs to be a commercial distributor?
I can remember back when I first installed Yggdrasil Linux on my 486 box in 1993 that the emacs install was like 80 megs. There was no way I could afford that much HD space on my tiny 330 meg hard drive.
BSD ports don't consist of source file sets that have been fucked with by the distribution assembler, like Red Hat SRPMs and what-not. The patches are instead applied during the build process to pristine tarballs of the source code from wherever the source originated.
The only XP computers 'suspectable to MSBlast' are those that are unpatched, which is significantly fewer than 'every single machine'. Just the same as the only Linux servers (known) to be susceptible are those without current patches applied. And there are plenty of Linux boxes out there that aren't kept up to date. Nobody told the people who put them online that the 'free' operating system was going to require 10 hours a week of reading Usenet security groups and update websites and applying patches.
Historically, Richard Stallman was one of the hackers at MIT who actively opposed the impostion of passwords on the Unix account logins. He and other hackers like him at the time opposed passwords because they believed in a community of sharing and openness. They refused to put passwords on their accounts for as long as possible.
Yep. And the same exemption-from-liability is common in the Free Software world. I think that right in the COPYING file there's an expressed 'no-liability' clause.
We'd better get right on that. I can see all sorts of free software developers needing liability insurance policies before they can push the new source tarball out on the ftp site.
Apple managed to do it by standing on the shoulders of giants,
Apple managed to do it by pissing away millions and millions of dollars trying to prove they were capable of producing a real OS with real multitasking, robustness, etc. All through the 90's they used up cute code-names on 'next-generation Operating System' buzzword sinkholes (object oriented this-n-that of the year).
Then they finally gave up and just put new greasepaint on NextStep instead. Yep. The NIHNI (not invented here? not interested!) brigade finally just went out and bought an OS because they couldn't roll one of their own.
Boy that had to hurt if you're a middle-aged old-line OS developer at Apple.
Maybe the firewalling should be done at the ISP. Block all incoming traffic to un-needed ports. If someone needs Port 80, open it up and bill accordingly.
Oh, wait. I hear all the screaming and fury coming from a mile away at that idea.
There isn't a heck of a lot of good engineering design software for Linux. There never will be in the form of Open Source. It's software that costs $2-30,000 per seat. You know, software for grownups, not dilletantes who browse the web and 'admin' common commodity tasks like web servers. We can't all just sell stuff and/or present it for sale. Somebody has to design it.
Microsoft is already on it. The whole .NET initiative. You didn't think they were going to keep on keeping on with the old ways, did you? Subscription software all the way. They'll make more money that way, and with them in control of your system, the problems seen here will go away, i.e. every one of the problem machines out there right now would already have been patched automatically.
This whole crisis is an opportunity for Microsoft to hype what they've been hyping recently (subscription software, downloaded from MS regularly).
It must have something to do with you having a hell of a lot of bandwidth available to download all that stuff over and over and over again with each x.x.x.01 revision. That won't scale very well to the whole world.
That day will never come. Enough of us are of an age to remember the days when there were fifteen different PC platforms out there and the huge splintered market for commercial software that resulted.
It's trouble enough for retailers to sell both Mac and PC games. Do you really think shrinkwrapped boxes are going to contain the seven CDs necessary to have the app run on 15 seperate OSes?
Yeah, everything will be distributed as source code. Uh-huh. People will like that.
Yep. And The Committee will decide what software you run on your PC. Better go get a Form # 34958/12 and fill it out. In a week or so (maybe) somebody from IT will roll out that executable and you'll be able to do your work.
Fuck that. It might work in a stodgy bureaucratic business enviroment, i.e. in big dinosaur corporate envionments. It will NEVER work in small dynamic companies, let ALONE in a home computing environment.