...but you couldn't prove it by looking at the majority of trade press out there related to computers. What you need is someone who is not in computers for the money first, someone who has ambitions of programming more than payroll or trading systems. An interest in C++ over Java or Python would be a good indicator. Linux over Windows. An interest in modelling the real world versus modelling money. I'd start with the computer science department at your local university, and follow leads from there.
The reason some of us find the BSD style licence inadequate is because it doesn't preserve free choice - I write software and release the source, you use the
freedom thus provided to make changes, and you sell the software (fine) but don't give your customer the source or the option to copy and redistribute (not
fine) - you take away their free choice.
But they had freedom of choice before they bought the proprietary software. They could have gone with your software. The only time they would lose this freedom is when a monopolist can compel usage of only their offerings in perpetuity. Thankfully that is against the law in at least the United States.
Right, I use "libertarian" as a substitute for "non-copyleft", i.e. BSD and X11 licenses. The definition of "libertarian" I'm using is the simple one, of allowing others as much freedom as possible. In general I believe in the right of an author to control their work, to publish or not to publish, to set the terms of reimbursement. But we are not talking about the rights of the original copyright holders with free software, we are talking about the rights of those programmers who make derivative works from free software.
If you start with libertarian (non-copyleft) free software, its seems you can use any license you want for a patch. These licenses don't restrict what you can do with derivative works, and patches are surely an embodiment of derivation. But the GPL gets into what you can do with derivative works, for better and for worse.
Who said the bedrock of open source includes anti-forking and anti-commercial sentiments? Dig deeper and you'll find the ideology of science beneath it all, an ideology that has no argument with independent derivative works or the accruing of private benefit from openly shared knowledge.
There are two avenues for open-source project communication on SourceForge: forums and mailing lists. Only the site-wide forums get any real traffic (the ones pointed to by the home page), and most projects rely on mailing lists instead. It is far easier and time-sensitive to receive periodic discourse in your e-mail inbox than having to log in to a site every day. I've disabled all but the developers forums on my SourceForge projects, and these are a special case, in that they are private, for the use of project developers only. Hard to imagine VA Linux getting into the business of republishing private forum content.
By the way, there has been some discourse at SourceForge on uniformly gatewaying the hosted mailing lists to a newsgroup hierarchy, similar to what Debian does with their mailing lists. No response on this idea yet from SourceForge personnel.
The one thing I haven't figured out yet is, if Computer Science (strictly) isn't really for software development, what is it for?
I find it useful for thinking about and discussing a wide variety of programming activity. Somehow studying the "pumping lemma" warps your mind in a way that makes recursion and self-reference easier to deal with. Like cutting up a dead body makes you a better physician. Without it you just don't have the same neural wiring. Unix and the Internet came from computer scientists, not software engineers.
If you want to start with the LGPL, you could consider adding a Stallman-approved binaries exception of the kind used by WxWindows. This preserves copyleft on all source distributions, but gives everyone the right to make proprietary binary products. If you want to reserve that right to yourself you'll probably have to wade in and try dual-licensing. Aladdin Ghostscript is the pioneer example.
You might also consider adding a copyleft term of some sort to the MIT (BSD without advert clause) license.
I don't like the idea of closed-source extensions to open protocols. Maybe someone knows how to add a clause to the BSD/MIT license to restrict this, I don't. A business person who practices this will do it whatever way they choose, which may or may not leverage BSD code. True, the GPL keeps some portion of open source out of their hands. I think it good protection for such central utilities as emacs and gcc.
But what about the development of custom applications for use in private enterprise in ways that are not general purpose or of interest to the general public? BSD/MIT allows an open source business to sell a service like that which can benefit free software and proprietary software at the same time. GPL protects the tools and OS. LGPL is ok for protecting general-purpose libraries. But BSD/MIT allows the enterprise to exist.
It is an alternative to earning a living working for speculative open-source companies building a brand around the GPL. How many can get venture backing for their business and win the IPO lottery? But everyone can try their hand at building custom applications.
I think your assertion that Red Hat is standard 'Linux'and your conclusion that Red Hat will become a monopoly, via network effects, is premature.
Perhaps. I would like to see diversity and competition take a permanent hold on computing, but the ideological and business forces that push for a single winner are still in place (at least in the United States).
the moment a more useful Linux distro appears, everyone who needs to jump ship will simply jump ship
Not if your employer/customer/co-workers require software delivered in the latest version of Brand X binary packaging, and only the latest version of Brand X binary packaging. Jumping between distros might be simple for a home user who wants to experiment with the latest one, but other things have to be taken into account in a work situation.
The idea that any open source code might be taken proprietary in a way that stymies the community is unfounded. A fork will instantly appear, like in the case of Xfree86 after the Open Group modified their copyright permission notice for X.
But don't forget the dreaded "network effects", the phenonemon that tends to restrict people's choice in the workplace to a single standard. I think it ironic that this new kind of dominant player arises in the computing world so shortly after we've all felt the relief of the former giant being reined in. But it makes sense that it would happen, since venture capitalists main criteria for investing in firms (and buying up firms) is what they call "unfair advantage". Usually it is patents or specialized know-how. But there is one opportunity in the open-source world to create such a behemoth, and RedHat/Cygnus is gunning for it.
It is doubtful in 15 years time RedHat/Cygnus will be broken up because they became a monopoly that retarded innovation in the computer world. I think they are smart enough to avoid the breaking the law. But they could have a lot of power over our computing future.
Here is a hypothetical scenario: what if RedHat/Cygnus makes the decision to focus their limited resources for backward compatibility testing of gcc to only that software found in the RedHat Linux distribution? And what if they decide to start deprecating features of certain languages (i.e. disallowing pre-standard C++). Do you think this would be fair? Do you think they might do it anyways?
As their position with respect to the open source community grows I will be watching how they continue to perform their important role of stewardship of the FSF's programming utilities. I second the comment someone made above of asking them to support gcc and gdb developers on the FSF's payroll instead of their own. Richard Stallman is the one monopolist I can put my trust in for this software, because he has more than earned it.
Slashdot Mirror
...but you couldn't prove it by looking at the majority of trade press out there related to computers. What you need is someone who is not in computers for the money first, someone who has ambitions of programming more than payroll or trading systems. An interest in C++ over Java or Python would be a good indicator. Linux over Windows. An interest in modelling the real world versus modelling money. I'd start with the computer science department at your local university, and follow leads from there.
But they had freedom of choice before they bought the proprietary software. They could have gone with your software. The only time they would lose this freedom is when a monopolist can compel usage of only their offerings in perpetuity. Thankfully that is against the law in at least the United States.
http://homepages.go.com/~martweb/AutoTrace.htm
and legal MP3s as well
Right, I use "libertarian" as a substitute for "non-copyleft", i.e. BSD and X11 licenses. The definition of "libertarian" I'm using is the simple one, of allowing others as much freedom as possible. In general I believe in the right of an author to control their work, to publish or not to publish, to set the terms of reimbursement. But we are not talking about the rights of the original copyright holders with free software, we are talking about the rights of those programmers who make derivative works from free software.
If you start with libertarian (non-copyleft) free software, its seems you can use any license you want for a patch. These licenses don't restrict what you can do with derivative works, and patches are surely an embodiment of derivation. But the GPL gets into what you can do with derivative works, for better and for worse.
Who said the bedrock of open source includes anti-forking and anti-commercial sentiments? Dig deeper and you'll find the ideology of science beneath it all, an ideology that has no argument with independent derivative works or the accruing of private benefit from openly shared knowledge.
By the way, there has been some discourse at SourceForge on uniformly gatewaying the hosted mailing lists to a newsgroup hierarchy, similar to what Debian does with their mailing lists. No response on this idea yet from SourceForge personnel.
I find it useful for thinking about and discussing a wide variety of programming activity. Somehow studying the "pumping lemma" warps your mind in a way that makes recursion and self-reference easier to deal with. Like cutting up a dead body makes you a better physician. Without it you just don't have the same neural wiring. Unix and the Internet came from computer scientists, not software engineers.
You might also consider adding a copyleft term of some sort to the MIT (BSD without advert clause) license.
Has the FSF accepted donations from Red Hat?
But what about the development of custom applications for use in private enterprise in ways that are not general purpose or of interest to the general public? BSD/MIT allows an open source business to sell a service like that which can benefit free software and proprietary software at the same time. GPL protects the tools and OS. LGPL is ok for protecting general-purpose libraries. But BSD/MIT allows the enterprise to exist.
It is an alternative to earning a living working for speculative open-source companies building a brand around the GPL. How many can get venture backing for their business and win the IPO lottery? But everyone can try their hand at building custom applications.
Perhaps. I would like to see diversity and competition take a permanent hold on computing, but the ideological and business forces that push for a single winner are still in place (at least in the United States).
the moment a more useful Linux distro appears, everyone who needs to jump ship will simply jump ship
Not if your employer/customer/co-workers require software delivered in the latest version of Brand X binary packaging, and only the latest version of Brand X binary packaging. Jumping between distros might be simple for a home user who wants to experiment with the latest one, but other things have to be taken into account in a work situation.
But don't forget the dreaded "network effects", the phenonemon that tends to restrict people's choice in the workplace to a single standard. I think it ironic that this new kind of dominant player arises in the computing world so shortly after we've all felt the relief of the former giant being reined in. But it makes sense that it would happen, since venture capitalists main criteria for investing in firms (and buying up firms) is what they call "unfair advantage". Usually it is patents or specialized know-how. But there is one opportunity in the open-source world to create such a behemoth, and RedHat/Cygnus is gunning for it.
It is doubtful in 15 years time RedHat/Cygnus will be broken up because they became a monopoly that retarded innovation in the computer world. I think they are smart enough to avoid the breaking the law. But they could have a lot of power over our computing future.
Here is a hypothetical scenario: what if RedHat/Cygnus makes the decision to focus their limited resources for backward compatibility testing of gcc to only that software found in the RedHat Linux distribution? And what if they decide to start deprecating features of certain languages (i.e. disallowing pre-standard C++). Do you think this would be fair? Do you think they might do it anyways?
As their position with respect to the open source community grows I will be watching how they continue to perform their important role of stewardship of the FSF's programming utilities. I second the comment someone made above of asking them to support gcc and gdb developers on the FSF's payroll instead of their own. Richard Stallman is the one monopolist I can put my trust in for this software, because he has more than earned it.