It's the same end, actually. Scare people away from publishing the material you want to sell at your amazingly inflated prices, pretend that it's incredibly valuable, and make sure they don't interfere with the money going to your top executives and not the people doing the hard work.
NVidia, at least, also replaces the OpenGL libraries with their own proprietary ones in order to fully enable their features. So it's not just hooking in a nonGPL kernel module. Even with the kernel bits published, it still won't work fully without those proprietary libraries.
They don't offer those to software purchasers. It's very difficult to get VMware running on new hardware platforms, partly because of this lack of source, and partly because their underlying kernel is so very old. Newer kernel patches simply can't be applied to it.
The likelihood of Microsoft doing it first is very low. The difficulty of finding and bringing the prior art to court, however, is very real, and encourages patent farming such as Microsoft engages in.
You claim it "wouldn't be too difficult". You've apparently never tried installing a set of 50 CPAN modules directly from CPAN, because some idiot added a dependency to some experimental bleeding edge new package that's never been tested in a setup with components as new as your other components. Or installing Oracle on anything. Or compiling gcc on a new architecture that doesn't have the glibc you expect.
It's a lot of work to build and test and verify good installers. I commend the Debian team for their work to date, and wouldn't want their job, because they have to deal with crack monkeys like Dan Bernstein who thinks his software is so special it should go in "/services" and not have any man pages.
In a world without copyright, if a vendor publishes the software and I reverse engineer it, modify it, or find a copy and republish it, I can act without fear of reprisal.
Then GPL isn't necessary, or mostly isn't. (Patents and trade secrets get tied up in it and get interesting to deal with.) As it stands, if I don't say "you have to share the code", people will take the code I write, walk away, and never let me see what they did, then sue me for copyright violation if I happen to write new code that is similar. And this is nearly inevitable with any large software project!
This can, and has, happened repeatedly. Take a look at the FSF pressure on router manufacturers who did exactly this sort of thing.
No one I met in the field called SunOS + Openwindows Solaris, until Solaris 2.x came out. We all called it SunOS, or SunOS 4.x when I was using it. Then Solaris 2.x came out, and all the Sun sales people called the older software SunOS and the newer release Solaris, until it became possible to close a sale by mislabeling the other version by the opposite name and the confusing release numbers.
The modern equivalent is the mislabeling of "RedHat Enterprise Release" as RHEL 1, which was really RedHat 10.0. Now that the release numbers have worked their way around to 5.0 again, my sesume of working with RedHat 3.0, the original one and not the newer RHEL 3.0, has become confusing.
I don't think "based on" is the right description for SunOS 4.x and Solaris 1.x. At the time, Sun claimed that Solaris 1.x was SunOS with Openview: they came in the same box. Similarly, Solaris 2.x was SunOS 5.x with Openview: again, they came in the same box.
The confusing labeling was part of a strange marketing attempt to confuse people into buying the new Solaris OS, but being able to say "it's SunOS" or "it's Solaris" depending whatever the sales manager needed to make the sale and confuse the customer into getting it. Most technical folks saw through this pretty quickly, but it made serious problems getting purchase orders written that would get the correct verson of the operating system. I caught Sun vendors flat-out lying when I asked for "SunOS 4.x" and they shipped "SunOS 5.x", and the hardware wouldn't run on SunOS 4.x. (This was an issue with the newer, cheaper, better sun4m hardware.)
This led to Tatung with a developer's license publishing a SunOS that ran on the new, cheap, very nice sun4m hardware and getting lots of hardware sales this way, and embarassing Sun into releasing a native SunOS that would run on it and killing their software sales business plans. They were really trying to discard what most of us call SunOS andn move to what most of us call Solaris.
They weren't real happy with the Sparc based Linux distributions, which took their fairly robust hardware and extended its life by many years with small Linux distributions, rather than paying expensive license fees for the next version of the software. Linux merely culminated Sun's irritation with open source projects like gcc, X11 with its forks, and gzip instead of their Sun compiler, Openview, and UNIX's compress tool.
It's off-topic. But NFS is a serious legacy problem. It's known as "No Fucking Security" for a set of strong reasons, and its "state-free" architecture has been causing file-locking problems for many years. The user authentication and access controls of NFS are awful, from a time when your hostnames were published via/etc/hosts and all the machines on your network ran exactly the same OS so had same usernames and uid's for everything..
If you need serious large-scale file sharing, use AFS. for shared common directories with no ocmplex symlink or access structures, use CIFS. Don't waste your time with NFS, because clients crashing can wildly confuse the server and screw it up even with NFS v4, which has never been easy to deploy in a large or mixed environment.
ZFS may server your needs well, I haven't had a chance to play with it.
SuSE. SuSe includes the Linux kernel and source code.
Novell has been publishing the Linux code themselves for years, especially since their purchase of SuSE, and contributing to it publicly in a way that SCO managed to somehow fail to admit for years with their Caldera product. Even attempting such a lawsuit would gut their SuSE business: on top of the wrath of the open source community over their patent deal with Microsoft, and the fact that GPLv3 is aimed square at that deal to prevent anyone from ever trying that stunt again, the anger of the open source developers would shoot a lot of their business through the head.
It would also make it hideously expensive to hire the good developers: they already lost Jeremy Allison, one of the core Samba contributors, when they did hte patent deal with Microsoft. They can't afford to lose other industry leaders this way.
SCO was about the FUD. The Solaris concern is reasonable, and the sort of things a few project managers and the lawyers should be able to resolve in a few meetings. They may even need to write up a few agreements to make sure that there are no grounds for future confusion. I agree, it's not a big problem.
There may be some serious features of SVRX to release under a public license and make available for develoopers. But I'd want it under GPL, not public domain, to protect its freedom to advance.
Because they don't. They hype a few popular artists, pay them ridiculous sums, destroy them with pressure, steal massively from them by paying "net" profits instead of gross and cooking the books to reflect ludicrous executive benefits and massive management overhead as "costs", and trash the careers of other new artists.
It's bad for the artists, it's bad for the art, it's bad for the consumer, and it's only good for the management of those few companies allowed to compete.
Oh, my. I'll assume this one is a troll, but answer it anyway.
Yes, they could release it under public domain, and it would instantly be grabbed, modified, and the modifications copyrighted so the GPL author could not use the variants, and would get sued for using those modifications. This is precisely what Microsoft tries to do with their "embrace and extend" approach to open source software, so it's not an imaginary problem. And it's what the early AT&T UNIX authors did, which is why Richard Stallman spent so much time reverse-engineering software in order to patch it, became understandably frustrated, and created the free software movement to preserve the freedome to write and modify software based on the ideas of other people willing to share their software.
If copyright did not exist, there would be little punishment available to people who don't want others able to modify, extend, or reuse their software, so the primary freedoms of the FSF would already be protected without slapping a GPL on it. OK, there are some remaining software patent issues, but the problem would be much smaller.
Keep in mind, though: copyright is not only about money for authors. It's about control of information. Copyright is used to prevent people from publishing the secrets of Scientology and other cults, it's usd to prevent publication of private correspondence, to prevent publication of internal government documents, prevent publishing modified copies of the Bible and the Quran that heretical followers might use, etc. Removing the money issue does not eliminate the need for or desire for copyright. We're seeing that right now in any place where the lawmakers want their secrets preserved, used to prevent publication of recordings of those lawmakers and their contributors.
It's a denial-of-service attack. By sending a few innocent sacrifical lambs with modified passports, it's possible to take out an entire security station's set of scanners and clob the entire security theatre operation of a small airport, and make fake passports much easier to get through the systems in the interim, especially if this becomes a common occurrence for a few days.
Better yet, modify the passports of people you don't like to delay their passages through security and get them arrested. There are plenty of other possibilities.
Oracle installers are notoriously bad, and seriously deform basic UNIX and Linux system configurataions. For example, "/a/b/c/d/.." is not the same as "/a/b/c". "dirname/a/b/c/d" is the same as "/a/b/c". And "cd/a/b/c/d; dirname `pwd`" is also not the same as "/a/b/c" in any system that uses autofs.
These are basics, but Oracle is not capable of doing them, and never has been. The result is that their software is not easily installed or integrated into any standard system that does not aggressively avoid practices common to UNIX and Linux systems administrators.
Re:And all of a sudden....
on
SCO Loses
·
· Score: 1
SCO has been extensively funded for the last few years by Microsoft-brokered "business partnerships". There are lots of articles on it, such as http://news.com.com/Fact+and+fiction+in+the+Micros oft-SCO+relationship/2100-7344_3-5450515.html. It's a good way for Microsoft to fund SCO and cast fear, uncertainty, and doubt on Linux software licensing while maintaining plausible deniability. Microsoft offers the other partners involved handsome licensiing terms or lucrative software fees to get them to deal with SCO.
If you've never been to one of those corporate software sales deal or partnership meetings, I highly recommend going, and taking really good notes on how much is done by reading between the lines andn by subtle "encouragement". It's fascinating to watch, and it explains a lot of corporate software purchases of mediocre or even bad software that serves some "business plan" that may not even come from your own company.
Microsoft cannot be safely seen to fund SCO directly, for lots of reasons. The nasty reaction of employees like yourself realizing just how evil their employer can be, as a matter of blatant fact, would be very expensive and would probably cause people like yourself to leave and be replaced by less skilled, untrained employees. And you can lose your best people this way, such as when Jeremy Allison resigned from Novell in the wake of the Novell/Microsoft patent deal.
You're working for a company that's been caught at far, far worse skullduggeries. This one is nasty, but minor by comparison.
More than just satisfaction: RedHat protected their own business, and faced a fraudulent litigator head-on. That kind of nerve to protect yourself and your customers is something software purchasers, like me, treasure in a vendor. They told the truth up front, they documented it, and held their own against the worst sort of lying weasel in court. It was very expensive: such lawyers are not cheap. Now they can focus their resources on more useful tasks.
Novell's success in the lawsuit against SCO buys them some credit in the open source community and respect from potential clients, such as me, that they badly need in the wake of their ill-managed decision to broker the weird patent protectons with Microsoft. They lost a huge, huge amount of support and credibility in the open source world. This will help them recoer some of it.
When Darl McBride goes to jail for fraud. And when SCO's books are laid bare and we can see *exactly* where SCO's budget came from for the last 5 years. It's clear from the quality and sales of their products that it hasn't been from selling software, and there are some fascinating reports of "partnerships" with third parties sponsored by Microsoft that kept them afloat while this lawsuit continued.
The object of the lawsuits from SCI have clearly been to damage the Linux community. I want to follow the money, or have someone competent like PJ over at groklaw.net follow the money, and see who was paying for the fraudulent lawsuits by SCO.
Seriously, atime is a performance hit and quite pointless except for a few legacy applications like mutt. It's amazingly bad for mail and news and web and build servers.
Is there any reason not to simply turn it off for/, besides mutt?
And to the people still thrilled by mutt and its use of easily parseable plain text files, seriously consider switching to a Maildir based system that provides the same unmodified email power, and doesn't have the historical limitations on maximum number of messages that mutt used to have and still may have.
Slashdot Mirror
It's the same end, actually. Scare people away from publishing the material you want to sell at your amazingly inflated prices, pretend that it's incredibly valuable, and make sure they don't interfere with the money going to your top executives and not the people doing the hard work.
NVidia, at least, also replaces the OpenGL libraries with their own proprietary ones in order to fully enable their features. So it's not just hooking in a nonGPL kernel module. Even with the kernel bits published, it still won't work fully without those proprietary libraries.
They don't offer those to software purchasers. It's very difficult to get VMware running on new hardware platforms, partly because of this lack of source, and partly because their underlying kernel is so very old. Newer kernel patches simply can't be applied to it.
The likelihood of Microsoft doing it first is very low. The difficulty of finding and bringing the prior art to court, however, is very real, and encourages patent farming such as Microsoft engages in.
The only reason to say that is so that you can sell your stock to suckers who believe you, or short-sell to anyone else that stupid.
You claim it "wouldn't be too difficult". You've apparently never tried installing a set of 50 CPAN modules directly from CPAN, because some idiot added a dependency to some experimental bleeding edge new package that's never been tested in a setup with components as new as your other components. Or installing Oracle on anything. Or compiling gcc on a new architecture that doesn't have the glibc you expect.
It's a lot of work to build and test and verify good installers. I commend the Debian team for their work to date, and wouldn't want their job, because they have to deal with crack monkeys like Dan Bernstein who thinks his software is so special it should go in "/services" and not have any man pages.
In a world without copyright, if a vendor publishes the software and I reverse engineer it, modify it, or find a copy and republish it, I can act without fear of reprisal.
Then GPL isn't necessary, or mostly isn't. (Patents and trade secrets get tied up in it and get interesting to deal with.) As it stands, if I don't say "you have to share the code", people will take the code I write, walk away, and never let me see what they did, then sue me for copyright violation if I happen to write new code that is similar. And this is nearly inevitable with any large software project!
This can, and has, happened repeatedly. Take a look at the FSF pressure on router manufacturers who did exactly this sort of thing.
Oracle installers being cross-platform is like the bird flu passing from humans to chickens. It's not really a good thing.
Thank you, I wasn't sure at this late date that the Solaris brand didn't exist until Solaris (2.x) came out.
It does seem kind of silly, doesn't it, to rebrand the older OS and leave the overlap of names in place? It even confused the Sun sales people.
No one I met in the field called SunOS + Openwindows Solaris, until Solaris 2.x came out. We all called it SunOS, or SunOS 4.x when I was using it. Then Solaris 2.x came out, and all the Sun sales people called the older software SunOS and the newer release Solaris, until it became possible to close a sale by mislabeling the other version by the opposite name and the confusing release numbers.
The modern equivalent is the mislabeling of "RedHat Enterprise Release" as RHEL 1, which was really RedHat 10.0. Now that the release numbers have worked their way around to 5.0 again, my sesume of working with RedHat 3.0, the original one and not the newer RHEL 3.0, has become confusing.
I don't think "based on" is the right description for SunOS 4.x and Solaris 1.x. At the time, Sun claimed that Solaris 1.x was SunOS with Openview: they came in the same box. Similarly, Solaris 2.x was SunOS 5.x with Openview: again, they came in the same box.
The confusing labeling was part of a strange marketing attempt to confuse people into buying the new Solaris OS, but being able to say "it's SunOS" or "it's Solaris" depending whatever the sales manager needed to make the sale and confuse the customer into getting it. Most technical folks saw through this pretty quickly, but it made serious problems getting purchase orders written that would get the correct verson of the operating system. I caught Sun vendors flat-out lying when I asked for "SunOS 4.x" and they shipped "SunOS 5.x", and the hardware wouldn't run on SunOS 4.x. (This was an issue with the newer, cheaper, better sun4m hardware.)
This led to Tatung with a developer's license publishing a SunOS that ran on the new, cheap, very nice sun4m hardware and getting lots of hardware sales this way, and embarassing Sun into releasing a native SunOS that would run on it and killing their software sales business plans. They were really trying to discard what most of us call SunOS andn move to what most of us call Solaris.
The man thought his mother was a virgin. Who knows what other confusions he suffered?
They weren't real happy with the Sparc based Linux distributions, which took their fairly robust hardware and extended its life by many years with small Linux distributions, rather than paying expensive license fees for the next version of the software. Linux merely culminated Sun's irritation with open source projects like gcc, X11 with its forks, and gzip instead of their Sun compiler, Openview, and UNIX's compress tool.
It's off-topic. But NFS is a serious legacy problem. It's known as "No Fucking Security" for a set of strong reasons, and its "state-free" architecture has been causing file-locking problems for many years. The user authentication and access controls of NFS are awful, from a time when your hostnames were published via /etc/hosts and all the machines on your network ran exactly the same OS so had same usernames and uid's for everything..
If you need serious large-scale file sharing, use AFS. for shared common directories with no ocmplex symlink or access structures, use CIFS. Don't waste your time with NFS, because clients crashing can wildly confuse the server and screw it up even with NFS v4, which has never been easy to deploy in a large or mixed environment.
ZFS may server your needs well, I haven't had a chance to play with it.
SuSE. SuSe includes the Linux kernel and source code.
Novell has been publishing the Linux code themselves for years, especially since their purchase of SuSE, and contributing to it publicly in a way that SCO managed to somehow fail to admit for years with their Caldera product. Even attempting such a lawsuit would gut their SuSE business: on top of the wrath of the open source community over their patent deal with Microsoft, and the fact that GPLv3 is aimed square at that deal to prevent anyone from ever trying that stunt again, the anger of the open source developers would shoot a lot of their business through the head.
It would also make it hideously expensive to hire the good developers: they already lost Jeremy Allison, one of the core Samba contributors, when they did hte patent deal with Microsoft. They can't afford to lose other industry leaders this way.
SCO was about the FUD. The Solaris concern is reasonable, and the sort of things a few project managers and the lawyers should be able to resolve in a few meetings. They may even need to write up a few agreements to make sure that there are no grounds for future confusion. I agree, it's not a big problem.
There may be some serious features of SVRX to release under a public license and make available for develoopers. But I'd want it under GPL, not public domain, to protect its freedom to advance.
Because they don't. They hype a few popular artists, pay them ridiculous sums, destroy them with pressure, steal massively from them by paying "net" profits instead of gross and cooking the books to reflect ludicrous executive benefits and massive management overhead as "costs", and trash the careers of other new artists.
It's bad for the artists, it's bad for the art, it's bad for the consumer, and it's only good for the management of those few companies allowed to compete.
Oh, my. I'll assume this one is a troll, but answer it anyway.
Yes, they could release it under public domain, and it would instantly be grabbed, modified, and the modifications copyrighted so the GPL author could not use the variants, and would get sued for using those modifications. This is precisely what Microsoft tries to do with their "embrace and extend" approach to open source software, so it's not an imaginary problem. And it's what the early AT&T UNIX authors did, which is why Richard Stallman spent so much time reverse-engineering software in order to patch it, became understandably frustrated, and created the free software movement to preserve the freedome to write and modify software based on the ideas of other people willing to share their software.
If copyright did not exist, there would be little punishment available to people who don't want others able to modify, extend, or reuse their software, so the primary freedoms of the FSF would already be protected without slapping a GPL on it. OK, there are some remaining software patent issues, but the problem would be much smaller.
Keep in mind, though: copyright is not only about money for authors. It's about control of information. Copyright is used to prevent people from publishing the secrets of Scientology and other cults, it's usd to prevent publication of private correspondence, to prevent publication of internal government documents, prevent publishing modified copies of the Bible and the Quran that heretical followers might use, etc. Removing the money issue does not eliminate the need for or desire for copyright. We're seeing that right now in any place where the lawmakers want their secrets preserved, used to prevent publication of recordings of those lawmakers and their contributors.
It's a denial-of-service attack. By sending a few innocent sacrifical lambs with modified passports, it's possible to take out an entire security station's set of scanners and clob the entire security theatre operation of a small airport, and make fake passports much easier to get through the systems in the interim, especially if this becomes a common occurrence for a few days.
Better yet, modify the passports of people you don't like to delay their passages through security and get them arrested. There are plenty of other possibilities.
NIH syndrome, also known as Not Invented Here.
/a/b/c/d" is the same as "/a/b/c". And "cd /a/b/c/d; dirname `pwd`" is also not the same as "/a/b/c" in any system that uses autofs.
Oracle installers are notoriously bad, and seriously deform basic UNIX and Linux system configurataions. For example, "/a/b/c/d/.." is not the same as "/a/b/c". "dirname
These are basics, but Oracle is not capable of doing them, and never has been. The result is that their software is not easily installed or integrated into any standard system that does not aggressively avoid practices common to UNIX and Linux systems administrators.
What?
Without copyright, the GPL wouldn't be necessary.
SCO has been extensively funded for the last few years by Microsoft-brokered "business partnerships". There are lots of articles on it, such as http://news.com.com/Fact+and+fiction+in+the+Micros oft-SCO+relationship/2100-7344_3-5450515.html. It's a good way for Microsoft to fund SCO and cast fear, uncertainty, and doubt on Linux software licensing while maintaining plausible deniability. Microsoft offers the other partners involved handsome licensiing terms or lucrative software fees to get them to deal with SCO.
If you've never been to one of those corporate software sales deal or partnership meetings, I highly recommend going, and taking really good notes on how much is done by reading between the lines andn by subtle "encouragement". It's fascinating to watch, and it explains a lot of corporate software purchases of mediocre or even bad software that serves some "business plan" that may not even come from your own company.
Microsoft cannot be safely seen to fund SCO directly, for lots of reasons. The nasty reaction of employees like yourself realizing just how evil their employer can be, as a matter of blatant fact, would be very expensive and would probably cause people like yourself to leave and be replaced by less skilled, untrained employees. And you can lose your best people this way, such as when Jeremy Allison resigned from Novell in the wake of the Novell/Microsoft patent deal.
You're working for a company that's been caught at far, far worse skullduggeries. This one is nasty, but minor by comparison.
More than just satisfaction: RedHat protected their own business, and faced a fraudulent litigator head-on. That kind of nerve to protect yourself and your customers is something software purchasers, like me, treasure in a vendor. They told the truth up front, they documented it, and held their own against the worst sort of lying weasel in court. It was very expensive: such lawyers are not cheap. Now they can focus their resources on more useful tasks.
Novell's success in the lawsuit against SCO buys them some credit in the open source community and respect from potential clients, such as me, that they badly need in the wake of their ill-managed decision to broker the weird patent protectons with Microsoft. They lost a huge, huge amount of support and credibility in the open source world. This will help them recoer some of it.
When Darl McBride goes to jail for fraud. And when SCO's books are laid bare and we can see *exactly* where SCO's budget came from for the last 5 years. It's clear from the quality and sales of their products that it hasn't been from selling software, and there are some fascinating reports of "partnerships" with third parties sponsored by Microsoft that kept them afloat while this lawsuit continued.
The object of the lawsuits from SCI have clearly been to damage the Linux community. I want to follow the money, or have someone competent like PJ over at groklaw.net follow the money, and see who was paying for the fraudulent lawsuits by SCO.
Seriously, atime is a performance hit and quite pointless except for a few legacy applications like mutt. It's amazingly bad for mail and news and web and build servers.
/, besides mutt?
Is there any reason not to simply turn it off for
And to the people still thrilled by mutt and its use of easily parseable plain text files, seriously consider switching to a Maildir based system that provides the same unmodified email power, and doesn't have the historical limitations on maximum number of messages that mutt used to have and still may have.