Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
GNU is Not Unix

A New Model for Software Innovation 317

Posted by michael from the out-with-the-old dept.
An anonymous reader writes "In this whitepaper published at LinuxDevices.com, Matt Asay (former Linux naysayer-turned-disciple) analyzes the GPL, picking apart what it means (and does not mean) for users, and whether it is enforceable. Assay also details how its terms inhibit and foster innovation, and why we should care. In this next generation of software, those who understand 'copyleft' licenses like the GPL will have the upper-hand, and will be best positioned to take on closed-source shops like Microsoft. Assay wrote this paper while attending Stanford Law School, where he studied the the GNU General Public License under Professor Larry Lessig." A thoughtful piece that answers - as well as they can be answered - a lot of the questions about the GPL that we get for Ask Slashdot, as well as examining the economics of it. Good reading for anyone developing or selling software.
This discussion has been archived. No new comments can be posted.

A New Model for Software Innovation More

A New Model for Software Innovation

Comments Filter:

  • Innovation (Score:3, Insightful)

    by bytesmythe ( 58644 ) <bytesmytheNO@SPAMgmail.com> on Wednesday August 28, 2002 @02:59PM (#4158224)
    From the article link (not the whitepaper):
    "In turn, this most 'anti-IP' of licenses is arguably doing more to foster innovation than patents or copyrights ever have."

    Patents and copyrights are about making money, NOT fostering innovation.

    • Re:Innovation (Score:3, Informative)

      by Milo Fungus ( 232863 )
      Interesting attitude. You need to read up on the history of copyright [harvard.edu]. It was originally intended "to promote the progress of science and useful arts." This is something that very few people understand, even many lawyers [slashdot.org]. The interpretation of copyright law has evolved into basically what you're describing, but that was not the original intent.
    • Patents and copyrights are about making money, NOT fostering innovation.

      Very true, but the excuse for patents and copyrights, and the economic and cultural harm such informational and intellectual monopolies cause, is the mistaken notion that they are intended to, or in fact do, "foster innovation." As we've seen demonstrated very clearly in the software industry, and as we've seen evidence of in other areas of endeavor, these government mandated monopolies do not foster innovation or progress ... so their very excuse for existing, and for granting such privelege to so few, is itself a farce.

      Which is something the average joe is only now, slowly, becoming aware of.

    • Re:Innovation (Score:4, Insightful)

      by Znork ( 31774 ) on Wednesday August 28, 2002 @03:15PM (#4158331)
      From the US constitution:

      The Congress shall have power ... to promote the progress of science and useful arts, by securing for limited times to authors and inventors the exclusive right to their respective writings and discoveries;

      The original intent of patents and copyright was about fostering innovation.

      Not that the corruption of the idea we have today is about that tho.

    • Making money and fostering innovation go hand-in-hand.

      (I would explain my position, but you didn't bother presenting anything that supported your statement, so why should I?)

    • Re:Innovation (Score:2, Insightful)

      by cmorriss ( 471077 )
      Patents and copyrights are about making money, NOT fostering innovation.

      I beg to differ. They are about making money AND fostering innovation. It is exactly the possibility of reaping large profits from a patented product that fosters innovation. The fact is, a vast majority of GPLed software is just a copy of a successful proprietary version. Some are good copies, but copies none the less.

      Is Mozilla more innovative than Internet Explorer or Opera?
      Is Linux (including the various desktops) more innovative than MacOS X?

      I would say they are not. In fact, I can think of MANY proprietary software programs that are truely innovative. The open source area produces few programs that fit this bill. They simply play follow the leader.

      • Re:Innovation (Score:3, Informative)

        by bytesmythe ( 58644 )
        Here's an excerpt from a talk given by Tim O'Reilly [oreilly.com]:

        "History teaches us that as far as innovation is concerned, open beats proprietary every time. You have only to look at the history of the UNIX operating system to see this effect. Many of the innovations that were incorporated into commercial UNIX systems (as well as many of the foundational technologies for the Internet) were developed in universities as extensions to the original work at Bell Labs. Once AT&T took UNIX commercial, under a restrictive license, that work stopped, and didn't burst into flower again until Linux, a free implementation, took over leadership of UNIX operating system development."

        When you close everything off, only enough innovation is done to perpetuate the monopoly, and no one else can share in the innovation to help further it.

      • Is Mozilla more innovative than Internet Explorer or Opera?

        Internet Explorer and Opera are based on the original NCSA Mosaic browsers -- which was an open source project. NT's TCP/IP Stack is mostly 'embraced and extended(corrupted)' BSD (open source) code.

        Without open source, Bill gates wouldn't have ben able to program his way out of a router.

      • Re:Innovation (Score:4, Insightful)

        by Stephen Samuel ( 106962 ) <samuel@NOsPaM.bcgreen.com> on Wednesday August 28, 2002 @05:58PM (#4159488) Homepage Journal
        Is Mozilla more innovative than Internet Explorer or Opera?

        Is TCP/IP and the Internet more innovative than xns? or netbios (which came much later)?

        Is Sendmail / SMTP more innovative than DEC's messaging system from the early '80s, or IBM's? (damned If I can remember their names).

        Is NFS more innovative than RFS?? (actually, I'd say that they were about even -- but NFS won out .. possibly because it was more open than RFS).

        ______

        When you think about it: VHS beat out Beta -- not because it was better but because it was more 'open' (just about anybody could produce a VHS machine, but only Sony could produce Beta).

        Similarly, many people ascribed DOS's ascendency over the obviously superior MAC to openness -- the fact that anybody could (and did) build a DOS box, and put whatever they wanted into it.

        Openness does drive innovation.
        With two products of equivalent inovativeness, the one that is more open tends to drive more collateral innovation -- and is thus most likely to thrive.

        Part of the reason for patents and Copyrights was the fear that, without them, big-money own-it-alls would usurp the works of truly innovative (usually small) authors and scientists -- make all of the money off of their work and, this remove any incentive for them to make their work known.

        Copyright and Patents have now swung to the other end of the pendulum. They've become and end in, and of, themselves. They've been 'strengthened' to the point where we'll soon be 'celebrating' a Century of Intellectual wilderness -- A century during which substantially nothing has been put into copyright that has subsequently flowed into the public domain -- as envisioned by the writers of the first amendment.

        Much of what was created in the early 1900s will be lost into this wilderness. Only the most famous (and, in some cases, the most mundane) works will survive. Much of the rest is caught in a legal limbo. Nobody knows who 'owns' it -- thus, nobody can obtain the right to copy it. By the time it's legal to redistribute it, the originals will be incapable of being duplicated.

        Consider that recordings of the technical conversations with the Apollo astronauts were only barely recoverable -- and this after less than 30 years in storage. Imagine what another 100+ years of languishing would do to less famous works -- Those who would like to preserve them risk criminalizing themselves in the process.

        Had the original works been protected under current Copyright rules, Disny's Snow White would never have been made. Similarly, most of Shakespeares works are known to have had their stories lifted from other authors.

        We build on the works of others. Copyright and Patent laws are useful to the extent to which they lubricate the path from creation to public availability. The moment that these laws block that path, they become repugnant to their moral and constitutional underpinning.

  • GPL Revision (Score:2, Interesting)

    by paladin_tom ( 533027 )

    The paper refers to a proposed revision of the GPL. A link to it is on the page.

    This GPL revision seems like a completely new GPL. It's language is much simpler, and devoid of legalese. Does anyone else worry about such a complete re-write being enforcable?

    • I don't like Matt's revision for one point - that it excuses things like dynamic linking as not being covered. Given today's COBRA and so on, any linking at all can be dynamic.
    • The current GPL is full of legalese because the FSF's lawyers go over it with a fine tooth comb. You're probably right that they wouldn't want to change the language.

      But it makes the GPL so much easier to understand, doesn't it? If the FSF doesn't like it, why doesn't the writer submit it to the OSI [opensource.org] as a new license? I'd use it.

  • While I welcome any insight or debate into the GPL and CopyLefting, frankly, I'm not getting anything new here. And the fact that he studied under Lawrence Lessig is going to imply some bias on his part, considering Lessig's positions. Frankly, other than speculation and name calling, there's not much real analysis out there on the issue of licensing in open source. What I'd like to see is some expert opposing viewpoints, starting a real debate on the legality of the issue. I've yet to see an article from the other side of the opinion fence regarding the GPL's legality, and this is strange, considering Microsoft's attack against the license, and the fear among many in the Open Source movement that it's unenforcable. How about it, Slashdot? How about some legal opinions from people other than Lessig or his students, pro AND con? That would be an interesting debate.

  • psych (Score:4, Interesting)

    by sstory ( 538486 ) on Wednesday August 28, 2002 @03:06PM (#4158278) Homepage
    Whatever the GPL says or omits, being 'best-positioned to take on closed source shops like msft' will not be possible without understanding why people will pay $550 for a word processor and spreadsheet. And dismissing that behavior as just 'lusers being stupid' isn't an answer.
    When people will pay $550 for an easy, intuitive word processor that comes with restrictions (Office), rather than a more primitive, ugly, yet free and somewhat as functional word processor (Emacs or other unix WP), there's something begging to be explained. And the licening details are of secondary importance.

  • lol (Score:2, Interesting)

    by Quasar1999 ( 520073 )
    former Linux naysayer-turned-disciple?

    That's funny... How come I don't read about former Microsoft naysayer-turned-disciple??? Anyone???

    • Re:lol (Score:2, Funny)

      by Anonymous Coward
      because they're too busy playing Sims and Warcraft III, right CmdrTaco?

  • Sounds great on paper (Score:3, Interesting)

    by Junks Jerzey ( 54586 ) on Wednesday August 28, 2002 @03:11PM (#4158307)
    I've read lots of papers like this, going back to Eric Raymond's rants. And you know, it sounds great on paper. It really does. And I love the idea of the GPL. But in practice, we're just not seeing all the innovation that you would expect. The classic examples are always cited--Perl, gcc, Emacs, Apache, the Linux kernel--but for a model that's supposed to be so superior, the success stories are fewer than you'd expect.

    One striking example is that the GPLed games for Linux always tend to variations of Tetris, Boulder Dash, Missile Command, etc. You would expect the innovative games to be coming out for Linux: no pressure from marketing, free development tools, big community. But the games with spark, like The Sims and Grand Theft Auto 3 are coming from elsewhere.

    • Re:Sounds great on paper (Score:4, Insightful)

      by Bruce Perens ( 3872 ) <bruce@perens.com> on Wednesday August 28, 2002 @03:26PM (#4158404) Homepage Journal
      The gamers are playing games instead of coding :-)

      The lure of Free Software development mostly has to do with useful software that can be developed incrementaly. It doesn't necessarily make much sense for games. The fact that both KDE and GNOME produced great desktops in 4 years is pretty impressive, given what folks have spent on CDE and so-on.

      Bruce

    • The classic examples are always cited--Perl, gcc, Emacs, Apache, the Linux kernel--but for a model that's supposed to be so superior, the success stories are fewer than you'd expect.

      And even those examples aren't really that pure. Perl is (arguably) the work one one loan, mad genius, and is not GPL.

      The gcc compiler is by far not the best compiler for any platform. It's sole advantage is that it's portable.

      Emacs is, well, Emacs. I'll refrain from commenting because of my personal biases. :)

      Apache is not GPL, so it doesn't reflect on the GPL, but it's probably the one application that is truly a success story. The reason, of course is that it filled a need when there was a huge demand for web servers and not many to fill the void. One can even make the argument that Apache represents the best web server, although you find plenty of argument about that in specific cases.

      The Linux kernel is certainly successful -- but far from innovative. No one (hopefully) claims that Linux represents even close to the "best" implementation of a Unix kernel. Still, you have to give props to the "usefulness" of Linux.

      • Perl is (arguably) the work of one loan, mad genius, and not GPL.

        But Perl is free software, and it is actually a lot more than the work of a loan, mad genius. It certainly started out that way, but the same is true of most good software.

        The gcc compiler is by far not the best compiler for any platform. It's sole advantage is that it's portable.

        Actually, on many platforms it's the *only* compiler. There were times where it generated the fastest code of any compiler out there for a few platforms, and I believe that is still true today on x86 (I could be wrong because I haven't benchmarked it against MSVC 7.0). In general, even proprietary compilers tend to not perform as well as the specialized, high-performance compilers that CPU vendors write for their own chips. It's true though that gcc's primary goals have been stability and portability. To that end they've been more successful than any of their competition in the proprietary space.

        I could go on. Heck, during most of the last 20 years and 90's it wasn't uncommon for admins to replace their vendor supplied Unix programs with GNU or BSD versions because they were better.

        Still, you have to give props to the "usefulness" of Linux.

        In general, free software doesn't tend to be terribly interesting from an academic standpoint. However, it does tend to be very "useful" (to paraphrase you). I'd argue that every piece of software you put on that list would score high on the usefulness scale. This is actually what you'd expect to be the emergent quality of software that is ultimately controlled by the end user. Don't undervalue "usefulness". That quality is by far the most important in software, and the one which is most lacking. If that were the only quality where free software exceeded proprietary software, it's reason alone to scrap the proprietary development model.
    • That's simply due to money!

      A grade "A" title single player game can cost $2-4M to develop. The majority of that figure is salaries for the people involved: Averages might be 2-5 programmers, 5-10 artists, and 3-10 support people, all over 18-24 months, plus overhead.

      Comments on game quality aside, most big games (lots of levels, high replayability, lots of art, lots of sound/music) take a heck of alot of money to develop. Someone has to front all that money. And if they think that releasing under the GPL will prevent them from recouping their investment, or have a significant negative impact on the bottom line, then it won't be released under the GPL. Period.

      What you have here is a simple incompatability of business models, or at least the perception thereof. Games traditionally have to make a pile of money fast, so investors can make a return, and salaries can be paid for the team to start on the second project. You're looking at $1-2M per team per year before overhead. How can a GPL model provide that?

      That's not to say it isn't theoritically possible to release the software under the GPL and sell all the content under another or traditional license, but someone needs to demonstrate that a GPL model can take in that $1-2M per year for a relatively short-lived product like a game before it'll happen.
      • Grade "A" operating systems can cost $100-1000M to develop. The majority of that figure is salaries for the people involved: Averages might be 10-1000 programmers, 10-100 artists and 30-100 support people, all over 18-48 months, plus overhead.

        [...]

        Oh wait, there's Linux, FreeBSD, etc, etc...

        I've got no idea why there are no Great Open Source Games With Spark, but I bet money's got nothing to do with it.
      • That's simply due to money!

        All of your points are correct, but I want to take it one step further: What's different about application development? Why is it possible to produce high-quality GPL applications under the same circumstances? In other words, is there anything we can learn from open-source app development and then apply to potential open-source game development?

        Here's where I'm coming from: I'm a professional game designer. I use Linux at home for (most) of my computing tasks, and that means I'm benefiting from the work of hundreds (thousands?) of open-source developers. I want to give something back, and I'm intrigued by the possibility of doing a GPL'ed, FREE game that isn't subject to the whims of the marketplace. That's where this thread's parent post starts to kick in -- once you get good, passionate people working on a game without having to worry about sales, suits, and the like, you're likely to get a riskier, more innovative result.

        Now... you don't need to tell me how long or how many people it takes to make good games; I know. But I can't help but think that there's a solution somewhere that we're ignoring. Again, what can open-source game developers learn from open-source application developers? Do applications really come together that much faster than games? Is it that there aren't enough artists willing to contribute their time and skills?

        (Afterthought: My post assumes that open-source games are being developed in volunteers' spare time; I'm under the impression that most open-source application development works this way. If this isn't the case, that changes the arguments.)
        • Its a tough problem, and I'm in the same boat. I'm a professional game programmer, about 15 years in the industry (SSI, Bethesda, Accolade, others...)

          10+ years ago I started an VR engine project that I wanted to eventually place into the public domain - I've even continued to get employment contractual waivers from employers to allow me to do so. Over the last 10 years I have invested hundreds of hours trying to cultivate relationships with others who could contribute.

          Here's the problems I keep running into:

          1a. High-end games are VERY hard, and most people who have the requisite skillset already work in games.
          1b. Most people who work in games have signed contracts precluding them from working on other games, including open source peojects.
          1c. People who could otherwise help are kept so busy with the sweatshop hours most game houses run on that they have little to no time available to contribute!

          2. If work on a game project is done on the side, it becomes very difficult to keep up with the state of the art. I've had to throw away huge sections of my codebase (especially 3d) and begin again to keep it competitive! It gets depressing when you know the code your working on at work is now n>1 years behind the engine you're building on the side. Low tech starting points do not inspire people to join...

          3. If development work on an open source engine is done in the open, it becomes trivial for other developers to "steal your thunder" by encorporating your best ideas into their titles, which due to funding, ship before you! It also becomes possible for certain parties to use their access to your work to begin patent abuse! (ala Rambus, etc.) Personal example: In the few weeks after I announced on my website that I was going to publish my VR scalability whitepapers, I had a huge traffic spike from the research.microsoft.com domain. I changed my mind...

          The other problem is related to tools. Games generally require significant content. Good content generally requires good tools, which generally means expensive. 3ds Max and Maya are very expensive. Artists who have purchased these tools for personal use NEED to make enough $ to recover their investment, not to mention keep food on the table. Audio tools are nearly as expensive...

          I think it could happen, but only if a sufficient critical mass was acheived. I haven't seen it happen yet...
    • ...games for Linux always tend to variations of Tetris, Boulder Dash, Missile Command, etc.[...]But the games with spark, like The Sims and Grand Theft Auto 3 are coming from elsewhere.

      Isn't it so that most OpenSource software is constructed to "scratch an itch". You have a need, you write a tool that helps you with that. You see someone who has a similar problem (or think that someone might have) so you give your tool away.
      You didn't write the program to make money, you wrote it to help you with a task.

      Games on the other hand is pure entertainment. You don't just sit down (most ppl don't anyway) and write a game. If you write a game, you probably want to make some mone off it. Especially if it's a new idea.

      .haeger

    • One striking example is that the GPLed games for Linux always tend to variations of Tetris, Boulder Dash, Missile Command, etc. You would expect the innovative games to be coming out for Linux: no pressure from marketing, free development tools, big community. But the games with spark, like The Sims and Grand Theft Auto 3 are coming from elsewhere.

      There's a very easy explanation behind this. Coding is one of the smaller parts (although certainly non trivial) in the effort involved in producing a top notch game. Finding quality artists who grok Open Source and are willing to open their works using an Open Source license is very difficult. This makes creating games like GTA3 and the Sims problematic.

      If you've ever gone in search of quality Open Source/Public Domain artwork/music you'll know what I'm talking about. The reason you see so many of these simple games produced by the open source community is that the coder is able to do the art themselves and therefore these are the games that get done.

      I would like to note however that free (as in speech and beer) art is getting more common as Linux has matured into a viable gaming platform. At WorldForge [worldforge.org] we're amassing large collections of music, 3d meshes/textures, sprites and sound fx in our CVS repository to help fix this problem. All of the material in our media modules are either Public Domain or GPL/FDLed (some older stuff is OPL). There is already a great deal of material in our cvs media repositories but much is still left to do before we'll see a vibrant indie gaming scene as you described above.

      -Jason
    • The examples you cite - Apache, Perl, etc - are also tools which have been lauded for their flexability under a variety of circumstances. It seems to me that once the major component is in place, innovation becomes a matter of what kind of modules you can tack onto a flexable system to cover a specific need in that system's baliwick. Therefore big success stories can be expected to be few and far between - a single flexable success story can cover a *lot* of ground, and the innovation then defers to figuring the niggly bits of how it applies specifically to you.

      As far as games go... well. I'd expect that the game programming going in is really done more in the spirit of "I wonder if I can figure out how they did that, then add my own touches to it", such as a CS student might employ to test their skills. These games have simple rules and a narrow environment. The sparky games like Sims and Grand Theft Auto shot for the moon in creating games with unique characteristics of gameplay and theme. The rules and environments are far more complex - "emulate various personality traits in a world full of objects that affect statistics in different ways, which then play off of the personality traits themselves to produce human-like behaviour... and make it skinnable" is a far cry from "blow up missles falling from the sky" or "push rocks down a dirt tunnel."

      Unfortunately, that takes a lot of talented and skilled developers working hard toward a declared, managed goal. The fact that these are ground-breaking ideas makes it all the more difficult and time-consuming. The sheer interactivity that *every object* in the sparky games carry with them is a thing of wonder to me, and I wouldn't expect anybody to pull something like that off in their copious spare time. The kind of game takes a fellowship of like-minded dedicated programmers who can put serious day-to-day time into the project -- and that means money.

      Not a programmer, just a guy who tries to think about things. I could be wrong.
      GMFTatsujin

    • You are looking in the wrong places (Score:5, Insightful)

      by FreeUser ( 11483 ) on Wednesday August 28, 2002 @03:51PM (#4158543)
      I've read lots of papers like this, going back to Eric Raymond's rants. And you know, it sounds great on paper. It really does. And I love the idea of the GPL...ut for a model that's supposed to be so superior, the success stories are fewer than you'd expect.

      That is because you are looking in the wrong places for your "success stories." You are wondering why some GPLed mapping program hasn't unseated Microsoft's Streetsmart, or why program X under the GPL so resembles commercial program Y (though often it is program Y which mimicks the GPLed program X ... much of OS X's eye candy can be traced back to the first program to ever use images in window decorations and borders: Rastorman's enlightenment, but I digress).

      What you are ignoring are the tens of thousands of small firms, like my own, like the dozen's a friend of mine consults for, and thousands of others around the world, who have used free software (GNU/Linux in particular) to build their infrastructure, their inhouse software, and their business out from under the everpresent heal of Microsoft, Sun, or Apple (though the latter has become a kinder master of late than the other two, it nevertheless remains a master). The success stories you are ignoring range from the traders I work for, who earn a sizable chunk of money thanks in no small part to GNU/Linux's superiority over their competitor's infrastructrue, to Pengiun Airlines, to Wallmart, to BMW, to the German government, the Chinese Government, the Brazilian government, and others too numerous to list, all of whome are able to run their IT infrastructure on their own terms, within their own budgets, and free from the coerced upgrades and chronic, forced obsolescence of their software and their hardware at the hands of their operating system vendor, as is chronically the case for anyone subject to Microsoft, Sun Microsystems, and yes, even kinder, gentler Apple.

      That is where you will find the innovation and the flourishing of the technology more than anywhere else, amongst the users which have been liberated by the Freedoms expounded upon by the Free Software Foundation, people like myself and many, many others who have been able to flourish even during these difficult times, thanks primarilly to the GPL and the user's freedoms which it helps to guarentee.

      Freedoms that allow us to run our businesses and earn a living, without being subject to the whims and deprivations of a copyright cartel or a convicted monopolist.
    • Makes me think...
      I really really beleive that the GPL is a brilliant concept.
      But, and sorry for that, I fail to see how I could buil a business on top of it.
      Sure, it can work for some type of application, you give the app, sell the support.
      But there are other type of application which I don't see could fit in this model.
      Games for one. How could you make a profit selling open-source games?
      First of all, there is no support to offer.
      Second, games are generaly obsolete within months.
      I mean, it is not like an open-source server which could evolve years after years.
      A game, you release it one time, maybe add to that a couple expension pack, a couple patch, and it's over.
      And as beautyfull an idea as it may be, and as much as I would like to share knowledge with each others, I have to eat, pay the bills and build up my future.
      All that takes real money!

      So, do you people beleive (and have a rational explanation) that the GPL can be applied to EVERY single type of software, or it can only be aimed at some categories of software which one can offer services for?

    • Re:Sounds great on paper (Score:4, Insightful)

      by pmz ( 462998 ) on Wednesday August 28, 2002 @04:44PM (#4158869) Homepage
      But in practice, we're just not seeing all the innovation that you would expect.

      What innovation are you expecting? Are you looking to be swept off your feet by the next handsome OS that you see in the elevator?

      Linux and GNOME are a unique twist on the old idea of a UNIX desktop. Linux has matured a great deal over the years to rival old giants like Microsoft, and GNOME has become a very good desktop environment for many people. While they aren't earth-shatteringly innovative, they provide some relief from the stagnant mainstream desktops out there.

      Then there are the countless tools that made my life livable over the years: GCC, gnuplot, ghostscript, Emacs, LaTeX, ispell, XFree86, Mozilla, and OpenOffice.org for my desktop, and sendmail, apache, and BIND on the servers. There are many many many more tools that I won't take the time to list out, but that doesn't mean they aren't there doing their part.

      Many Open Source software projects are not truly innovative but provide best-of-breed implementations of standard protocols. Nearly all of the standard Internet protocols have been implemented for Linux and the BSDs. Ones like Apache steal the show, and ones like the BSD TCP/IP stack are "borrowed" by smart and successful people like Microsoft.

      Perhaps the biggest contribution of Open Source is properly moving things into the "public domain" after no longer being commercially worthwhile or after becoming commodities. The protocol implementations I mentioned certainly fall into this category. For example, I seriously think the idea of a commercial web browser is obselete, and ones like Internet Explorer really serve to hold back progress rather than help it.

      Another way that Open Source contributes in ways that companies can't is to provide uncorrupted software. By "corrupted" I mean: integrated DRM, proprietary file formats, proprietary communications protocols, unfair EULAs, and any other blatant company or government-driven attempt to limit what computers can do. The frequent advocacy by Open Source enthusiasts for open file formats is one way to drive innovation that really can be earth-shattering (imagine a world without .doc!).

      So, I'm not sure what you're looking for, but I'm pretty convinced that there is more innovation occuring than we can shake a stick at. It isn't always suprising, revolutionary, or even obvious, but to say it isn't happening is to be in denial.
    • I've read lots of papers like this, going back to Eric Raymond's rants. And you know, it sounds great on paper. It really does. And I love the idea of the GPL. But in practice, we're just not seeing all the innovation that you would expect. The classic examples are always cited--Perl, gcc, Emacs, Apache, the Linux kernel--but for a model that's supposed to be so superior, the success stories are fewer than you'd expect.

      Fewer than you would expect? If by "success" you mean "market penetration", yes, we have problems. But we also lack dedicated marketing departments. If by success you mean "large, useful, stable, worthwhile", we're doing great. The Wine project is making amazing progress on effectively reimplementing Windows. It's to the point where Microsoft Office runs. Gnumeric isn't perfect, but for 90% of the world's spreadsheet needs, it's great. AbiWord and KWord are lagging a bit, but making great progress. I've been using GnuCash for almost a year and haven't missed MS Money in the slightest. Sure, the Gimp has some issues that make it difficult to use for professional image publishing, but it's more than adaquete for your average person, or someone doing web images. (A friend who does photo editing work for a local newspaper says that the Gimp has everything she needs to replace Photoshop.) Playing music: XMMS. DVDs, DivX:-), and many other video format: Xine. Email: Evolution. Highly standards compliant web browser: Mozilla. Every single one of these is a huge success story to me, they've allowed me to replace Windows as my primary work system without missing anything.

      One striking example is that the GPLed games for Linux always tend to variations of Tetris, Boulder Dash, Missile Command, etc. You would expect the innovative games to be coming out for Linux: no pressure from marketing, free development tools, big community. But the games with spark, like The Sims and Grand Theft Auto 3 are coming from elsewhere.

      There are a number of reasons why Free Software games tend to be more simplistic. The biggest is art, sound, and music. Programmers can whip up simple icons for their word processor, but generating the hundreds of textures, dozens of sounds, and music for a game isn't something we're as good at. And the concept of "Free" hasn't infected the artists and musicians to the extent that it has programmers. (This may be because programmers have a great deal of incentive to share code. It's harder to take another artist's work and tweak it a bit for your own needs.)

      That said, Free Software games aren't totally dead, just small. Nethack avoids the art problem entirely and remains one of the most innovative computer hack and slash games available. Freeciv is highly derivative, but has an innovating client-server set up. Tools like SCUMMVM and MAME show a great deal of innovation in reverse engineering and improved visual displays of existing content. Crystal Space is an excellent 3D engine mostly suffering from lack of content created with it.



    • One striking example is that the GPLed games for Linux always tend to variations of Tetris, Boulder Dash, Missile Command, etc. You would expect the innovative games to be coming out for Linux: no pressure from marketing, free development tools, big community. But the games with spark, like The Sims and Grand Theft Auto 3 are coming from elsewhere.

      As others have pointed out, modern games are becoming more and more a blend of technical and multimedia artistry. Part of this may simply be exposure and time will bring more artists in to the Open Source community. But some of it is also the culture and nature of artistic work.

      On the technical side, we already have some great coders working on Open Source tools. Part of this may be that coders are the first to benefit from other Open Source projects and feel compelled to "give back" to the community. But also part of this is definitely the positive return from finding others to collaborate on your project. Development of tools benefit from multiple developers. It might be worth noting that there are some very interesting Open Source projects involving toolsets and engines for games.

      The artistic side comes from a different world. First, most of their tools are proprietary... although with tools like the GIMP and Blender (knock on wood), this may change. Secondly, art tends to have a limited collaborative nature. And finally, art does not need an equivalent of source code to improve from one example to another. In fact, artistic influence is largely unaffected by IP laws. For example, while reading many of the better fantasy novels available today... one will often see the influence of J.R.R. Tolkein - despite a rather aggressive protection of copyright by the Tolkein estate (just ask TSR/WotC or whatever they're called now). This artistic influence is felt within other aspects of artistic endeavor also. Simply put, artists do not have as much a need for Open Source ideals as coders do (although "look and feel" and other issues are very similar technical examples of influence).

      But that doesn't mean artists will never find themselves in the Open Source fold. As more and more interesting tools begin to appear, Open Source may begin providing an environment that is attractive to artists' desire to experiment. And this experimentation is often produces some of the most compelling content.

      It may simply be a mater of time.

  • Innovation and interest (Score:5, Insightful)

    by Bruce Perens ( 3872 ) <bruce@perens.com> on Wednesday August 28, 2002 @03:12PM (#4158312) Homepage Journal
    There's a quote of Lincoln on one side of the patent office building in Washington DC. It says something about the patent system coupling innovation with interest, which I guess means providing a reward for innovation. That's where the money comes in.

    But of course there are any number of situations where that reward is not the optimal means of promoting innovation, or isn't even appropriate - one example is the case of public-funded research resulting in private patents, another would be when the monopoly grants not merely methods but entire categories of business, as with today's broadly-interpreted business method patents.

    Bruce

    • Re: (Score:3, Informative)

      by account_deleted ( 4530225 )
      Comment removed based on user account deletion

  • IP (Score:5, Interesting)

    by sstory ( 538486 ) on Wednesday August 28, 2002 @03:15PM (#4158328) Homepage
    regarding the ip/anti-ip content, this is a very old argument. As old as the american copyright system.

    Ben Franklin: I'll make my inventions free so humanity can reap the rewards.
    Tommy Jefferson: We'll make a patent system so that profiting from IP will promote invention.

    Because there's not overwhelming evidence that one or the other positions is superior, it usually becomes faith-based reasoning--the worst kind.

    • Re:IP (Score:3, Informative)

      by Jack William Bell ( 84469 )
      Normally I would ignore this, but Thomas Jefferson is one of my heros (Ben Franlin another) and I should point out that he was originally *against* the patent system and was never very happy with it. He also felt that humanity should be free to reap the rewards of invention. He worked very hard to maintain strict limits on patent terms and to limit just what *was* patentable.

      In fact he accepted the position of first head of the patent system precisely because he wanted to make certain nothing was patented that wasn't truly innovative and new.

      Please take more care when quoting, or paraphrasing. Anal retentives like myself are watching...

      Jack William Bell

    • Re:IP (Score:2)

      by dirk ( 87083 )
      regarding the ip/anti-ip content, this is a very old argument. As old as the american copyright system.
      Ben Franklin: I'll make my inventions free so humanity can reap the rewards.
      Tommy Jefferson: We'll make a patent system so that profiting from IP will promote invention.

      Because there's not overwhelming evidence that one or the other positions is superior, it usually becomes faith-based reasoning--the worst kind.

      While back then and even today on a basic scale there is no way to tell which one works better, on a giant scale such as today's R&D departments, it should be fairly clear that limited patents will encourage more innovation. It's a simple matter of economics. If it takes 8 million dollars (a fairly low price) to research a product and bring it to market (with no guarantee that there will be any return on the investment) very few people will take that risk knowing that everyone can simply copy their idea immediately and sell the product at a much lower cost, since they don't have to recoup R&D costs (not to mention R&D costs for products that failed). There would still be people trying to create new products, but there would be significantly less of them, because it could only be done by people/companies that don't plan on making money. If there is no patent protection, the person shelling out the money to innovate will always be undercut by the guy who takes his idea and can sell it for less.
  • I see this as the perfect chance to plug openchallenge [openchallenge.org] and maybe get good comments on how to improve the concept.

    Mace Moneta [67.80.55.197] summarized the idea well:
    " get those that have the skills needed to tackle a problem together with those that have the problem, and solve it as Open Source. The OpenChallenge web site [openchallenge.org] lets you submit problems; those looking for a project to work on can browse the list of requests for something that they find interesting. "

  • Either that or a robot. On page 17 he goes off on the "GNU/Linux" thing without provocation. Unless you consider the word "Linux" provocation. Anyhow, I just had to laugh. I guess I should have seen it coming but the FSF guy replied to to question in such a normal way and RMS had been replying so normally that it caught me off-guard.
  • Assay also details how its terms ...

    It's a great esay written by Assay.

  • proposed revision of the GPL (Score:5, Interesting)

    by Dr. Awktagon ( 233360 ) on Wednesday August 28, 2002 @03:21PM (#4158375) Homepage

    An interesting streamlining of the GPL. However he takes out all the "social statements" which may or may not be a good thing.

    A few comments:

    You can use GNU GPL Software for any purpose. You can modify, copy and distribute GNU GPL Software, including derivatives, for commercial or non-commercial purposes.
    In return, you must agree:

    No no, you can use the Software for any purpose. Period. If you re-distribute the software, THEN you must agree to the terms.

    The license should emphasize that it only covers "copying, distribution and modification". No one should feeled compelled to accept the GPL, just because they are using the software. This also helps dispel the stupid notion that you don't have a right to use software you buy or download unless someone grants it to you.

    This, in my opinion, is one of the greatest thing about the GPL. It lies dormant and only applies to you if you distribute copies. You don't need to read it, you don't need to "click-through" it. You only need to care when distributing copies, which you aren't allowed to do to begin with. In fact, if I were writing the GPL, I'd put that up at the very beginning (after a warranty disclaimer if that's necessary): "you don't need to read this unless distributing copies" or something similar.

    if you distribute GNU GPL Software, you include two things: a copy of this license and a ready means of obtaining the source code for the Software.

    Define "ready means"..

    That if you distribute a derivative of GNU GPL Software, you include a notice explaining the changes you made.

    (Nitpick) why isn't this included with the previous requirement? What constitutes a "notice explaining the change"? A diff? A paragraph? Is this important?

    The GPL is wordy, partly because it spells out some of this stuff in more detail (for instance, see item #3 in it). The "revised GPL" should also be specific whenever possible.

    Anyway, the GPL does need to be streamlined and clarified to a certain extent, and perhaps the FSF can get some ideas from this revision. For instance, cutting out the Preamble of the GPL entirely might be a good idea, and replace it with a link to the FSF philosophies.

    • Unfortunately, it may still be necessary to agree to the GPL even to use software. It frees the developer from liabilities by disclaiming all warranties including fitness for any purpose.

      • The question is whether or not a simply warranty disclaimer requires agreement. I'm not sure that Sprecht v. Netscape (on the nature of agreement) is relevant to that. Now, if you want the user to indemnify you, I would think that really does require agreement.

        Thanks

        Bruce

    • Re:proposed revision of the GPL (Score:4, Insightful)

      by Bruce Perens ( 3872 ) <bruce@perens.com> on Wednesday August 28, 2002 @03:35PM (#4158464) Homepage Journal
      You left out the part about linking. If we are to clear up that part, I'd much prefer to make it clear that some forms of dynamic linking and server-izing (as with CORBA) do create derivative works. I think the key concept of a derivative work in software has nothing to do with linking, and everything to do with the span of the thread of control in a computer program. This can indeed happen across physical system boundaries. The boundary for the copyright should probably be at an API that is explicitly designated as an interface to separate works, such as the kernel system call interface. Not just any API.

      Thanks

      Bruce

      • The licence cannot define what creates a derivative work. That is given by the copyright law and court decisions. Your arguments about where copyright's boundary should lie make sense, but they don't belong in the GPL's text.

        Of course the GPL could waive rights: 'even if for some weird reason it turns out that mentioning the name of Richard Stallman makes all future programs you write a derived work of GNU Emacs, we choose not to restrict distributions of such derived works'. But you are talking about strengthening, attempting to claim control over things which the law might not consider to be derived works, and that isn't possible. (Without some unpleasant UCITA/click-through which I'm sure RMS wants to avoid, and probably not even then.)

        ('IANAL' considered redundant.)
        • The licence cannot define what creates a derivative work.

          It could make more of an attempt at this if it purported to be a contract, but it does not. However, copyright law is fuzzy enough about what constitutes a derivative work in software - the law says almost nothing about it - that a statement included in the license of what the parties consider to be derivative works might have some force.

          Bruce

      • Well stated.

        The proposed revision of the GPL does "fix" some of the murkiness related to dynamic linking -- but does so by implicitly claiming them as derivative works.

        I know this is what RMS wants, and perhaps it's what the community as a whole wants, but it's also why most commercial companies won't touch GPL code with a 10' pole.

        At the very least an API as you define it should be able to form a firewall between GPL and non-GPL code (note - not necessarily commercial - I may want to distribute my code under BSD license, but it still needs to be untainted by GPL for that to be an option).

        Without this provision companies will continue to avoid GPL code for fear of losing core business logic. I code in Unix, use open source tools, and we avoid any GPL libraries even though there's no intent to distribute the code outside our company. It's simply not worth the risk that one day we'll have a component that does need to be distributed and potentially tainting everything else -- which is core business logic.
        • it's also why most commercial companies won't touch GPL code with a 10' pole

          Both HP and IBM touch the GPL with some simple policies and a review board as their only precautions. They tell their customers about it, too. It happens that I am in the business of selling corporate Open Source policy and practices, so I can say sincerely that the 10-foot thing is ending.

          People who want to save their business logic put it in non-GPL applications and don't put GPL code in those applications. It's not such a big deal. Also, if you do end up infringing, your proprietary code isn't "tainted", you have to cure the infringement. Writing out the GPL code and saying "I'm sorry" is an option to cure the infringement. Do you really think any court is going to compel someone to GPL their application?

          Bruce

          • It happens that I am in the business of selling corporate Open Source policy and practices,

            Somebody I know was recently hassled by a telemarketer trying to sell her aluminum siding for her house.

            She told the salesman: "But our house is made of brick!"

            The salesman answere: "We can put aluminium siding on a brick building....."
      • The boundary for the copyright should probably be at an API that is explicitly designated as an interface to separate works.

        This thought is good/interesting. Comparing the linking interpretation with the above, I conceive of the scenarios this way:

        (a) Linking: if you link to GPL'd code, you must release the entire source base

        (b) API Boundary: if you modify code within the formally designated API boundary, you must release your modifications within that boundary

        (b) seems more in the spirit of what I'd want, but it does present some problems:

        1) what rules are adopted to ensure that the new API boundary remains clearly delineated within the new additional code, with no holes or ambiguities?

        2) what if the person modifying the code decides to specify the API boundary at such a shallow depth into their added code that the spirit of the standard is defied, and the released code ends up getting polluted with such intentionally devious boundaries?

        .

        • Well, the big question is whether or not someone can arbitrarily add new API boundaries to your GPL code - as in your #2. Perhaps there should be some rules about the purpose of the API? I haven't thought it out, and I see this needs thinking through.

          Thanks

          Bruce

      • I'd much prefer to make it clear that some forms of dynamic linking and server-izing (as with CORBA) do create derivative works. I think the key concept of a derivative work in software has nothing to do with linking, and everything to do with the span of the thread of control in a computer program.
        However, this moves a major burden to the USER of the software, who is supposed to be free from restrictions on the use of the GPL-licensed software.

        By including dynamic linking (or CORBA) into the definition of derived works, the user must be aware of all code that may be imported into the application. A GPL program may be legal to use on one system (a linux system with a GPL'd corba instance) but not on another (a linux system with a GPL-incompatible corba system). Even if these corba implementations are indistinguishable from the application's point of view.

        Another example, would installing a non-free libc.a (with the same function prototypes, but not a derived work of GNU libc) on a linux system violate the license of all GPL software on that system? Any GPLd program which (dynamically) links with /usr/lib/libc.a would be combined (in memory at least) with something that can not be distributed under the GPL.

        --Joe

    • It's only a proposal, not the final version! Don't hammer it, it's not even complete yet.
    • For instance, cutting out the Preamble of the GPL entirely might be a good idea, and replace it with a link to the FSF philosophies.

      No, that would be a bad idea. The preamble is part of the licence, and would be used in the licence's interpretation in court.

  • ...will be on the Slashdot effect.
  • This guy deserves a new acronym

    IANALBISL

    I Am Not A Lawyer, But I Study Law.

    -Jon
  • I have just skimmed his paper (interesting stuff to me started on page 31), so this is tentative: I have been thinking of GPLing my software technology for semantic extraction and natural language processing.

    My problem is that I have invested a little under 2 years of work on my NLP tools, and although I only make about $5K a year selling them, that money does help pay the bills and I believe the long term market for my software is excellent.

    I have received many great bug fixes and code improvements for my Open Source projects and even more great help with material for my free web books, but these projects have not cost me nearly two years of work.

    I guess that the real question is if companies (and perhaps individuals) who use GPLed software to make money in their own closed commercial products are honest about paying a fair license fee to get a non-GPL license to use my work. I do receive many generous donations for my free web books, so I do trust in many people's generousity and fairness.

    An alternative to using the GPL is to license software for free non-commercial use and then hope that commercial use users act fairly.

    I think that many developers would very much like to have their work freely used by people who make no money off of their work and at the same time collect reasonable payments from anyone making money on one's work. Tough problem - please share any good ideas :-)

    In any case, it is tough for me to make a decision here.

    -Mark

  • One item only partly addressed in the document has been the increasingly burdensome requirements of Microsoft licenses (product keys in manuals, product keys stuck to the back of PCs - good idea MS!, online product activation and shotgun EULAs giving MS the "right" to alter your system). In combination with the increased cost of licensing, this is prodding all Windows consumers to re-evaluate their options. Not to mention MS's famed security...

  • The Author Does Not Understand the GPL! (Score:5, Insightful)

    by Gorobei ( 127755 ) on Wednesday August 28, 2002 @04:12PM (#4158689)
    He's confused about licenses versus copyright, e.g:

    The GPL's language, then, tries to swallow every piece of software that incorporates a GPL program or portions of it. However, it probably cannot do so under the copyright law it purports to follow. The definition of a derivative work proves highly controversial in the software world, and may not stand up in court. [Page 15.]

    This has nothing to do with the GPL!

    If I create a derivative work from a copyrighted piece of GPL software, I can release it under my own terms without accepting the GPL. I then probably get sued by the COPYRIGHT holder, and the courts then decide if this was a derivative work, a new creation, a copyright infridgement, or whatever.

    Alternatively, I can accept the license (the GPL,) in which case copyright law no longer applies: I have freely entered into a contract that gives me certain benefits and takes away certain others (e.g. my right to claim copyright trumps the license I just agreed to.) If I now start doing evil like violating the GPL, I get sued for violating a LICENSE, not copyright law.

    The GPL does not "build on" copyright law - it's a license that I can agreed to or not. If I don't agree, copyright is the operative thing, if I do agree, the license if the operative thing!

    • by
      modifying or distributing the Program (or any work based on the
      Program), you indicate your acceptance of this License to do so, and
      all its terms and conditions for copying, distributing or modifying
      the Program or works based on it.

      Seems pretty clear to me. The act of modifying the program, or distributing it, and the act of agreeing to the license are one and the same. In other words, you can do what you like with the software, but if you distribute or modify it, then you have already agreed to be bound by the terms of the GPL. You might imagine that you hadn't agreed to the license, but how else would a court determine whether you had agreed to it or not ?

      • You might imagine that you hadn't agreed to the license, but how else would a court determine whether you had agreed to it or not ?


        If you didn't agree to the licence, you have no legal means of distributing the code - you would be breaking copyright law.
      • The usually "defenses" by the GPL-violators appear to be:

        1) A rogue summer-programmer copied the GPLed code into our product. [This probably doesn't fly - he was their agent.]

        2) We re-implemented the code (fair enough, though most "reimplementations" are just rearranging and changing variable names, which probably will not stand up in court.)

        3) We never read the GPL, so we couldn't have agreed to it. E.g. we found the routine on the web using a search engine, cut&paste, didn't read the rest of the file.)

        4) We based our program on a stripped copy of the source we found somewhere - it contained no GPL.

        5) The GPL is weaker than a click-through licence, we didn't even click to accept!

        2 is completely a copyright issue.
        3 & 4, if found credible by the court, may make the whole thing come down to copyright.
        5 is interesting, but I don't know enough to judge.

  • Please don't read it (Score:4, Insightful)

    by T.E.D. ( 34228 ) on Wednesday August 28, 2002 @04:16PM (#4158720)
    Normally we have trouble with folks posting here without even bothering to read the article in question. But in this case I'd argue that would be a good thing.

    I read through about the first half of it, but his explanations of the GPL are so confused and horribly unsound that attempting to follow it was starting to muddle my brain.

    In the first paragraph it struck me as odd that he claims that he was recently convinced of the GPL's innovation harming qualities, but now is convinced that the complete opposite is true. Executive summary: I know I sounded convinced yesterday. But I was actually full of shit then, and just didn't realize it. But today I really do know what I'm talking about. Honestly

    However, after reading his embarassing attempts at logic, its perfectly understandable how something like that could happen to him. With logic that muddled, tomorrow he might be convinced the GPL has manged to retroactively kill the dinosarus.

    My suggestion is that anyone who doesn't have a real good handle on the GPL already avoid this whitepaper like the plague. Folks who fully understand the GPL should take the Monty-Python "World's Funniest Joke" approach, and split up into teams with each person analyzing a small bit of it seperately. That way perhaps no one person will read enough of it to become fatally confused.
    • I agree. The article was bad news. What the article fails to understand is that what my opinion or this guy's views of the GPL is is irrelevant. It is the opinion of a judge someday to decide the fate of Intellectual Property that links or otherwises uses GPL (and even LGPL) code. This is the fear that companies face with the GPL, uncertainty.
  • Let's see ... on the page of the paper, the author asserts that the public and corporate perception of Linux has undergone a 180 degree reversal in the last year. (Simply not true ... a year, even two or three years ago, Linux was very much on most of the mentioned companies' radar. He's trying to pass off his change of mind by claiming that it was the industry's) He also states that "Linux and its global network of developers have long sought nothing less than 'total world domination.'" (Again, untrue; 'Linux,' as an operating system, doesn't have any desires. Linus Torvalds, the head of the "global network of developers," has repeatedly stated that Linux is a hobby, not a grand political scheme. The only one interested in world domination is RMS, however, in that case the article shouldn't be about Linux at all; it should be about GNU and the FSF.)


    I'd take any paper that starts that way with a grain of salt.

  • how 'bout a little puppet show (Score:3, Funny)

    by Eil ( 82413 ) on Wednesday August 28, 2002 @04:34PM (#4158816) Homepage Journal

    <RMS MODE ON>

    Any part of Mr. Asay's article where he implies the GPL is bad or could use some work is merely due to his failing to comprehend or understand the meaning and scope of what the GPL in its entirety.

    <RMS MODE OFF>

    Isn't that pretty much what we're going to hear in a few days? I'll thank the moderators for at least attempting to have something resembling a sense of humor.

    P.S. No offense meant to RMS or his devout apostles.
    • Any part of Mr. Asay's article where he implies the GPL is bad or could use some work is merely due to his failing to comprehend or understand the meaning and scope of what the GPL in its entirety.


      Actually, it would be more accurate to say "Any part of Mr. Asay's article where he implies anything whatsover about the GPL, good or bad, is merely due to his failing to comprehend or understand..."

      Or better yet, borrow a disclaimer from TV:
      "This whitepaper is a work of fiction. Any resemblance to actual facts found in this paper, either living or dead, is purely coincidental."

      You'd honestly be better off learning about the GPL from reading Microsoft whitepapers. That's how bad it is. At least Microsoft will use decent logic after spreading the initial foundation of mischaractarizations. This paper is about as accurate and well-reasoned as a statement from Lyndon Larouche.

      I can tell you didn't actually read it before posting, but in this case I won't gripe at you for that, because on the balance that's probably a good thing. I read about halfway through, and just now got back from the subsequent bathroom trip (I feel much better now).

  • An FSF Response to Asay's article (Score:3, Informative)

    by bkuhn ( 41121 ) on Wednesday August 28, 2002 @04:58PM (#4158987) Homepage
    We at FSF have spent a few minutes reviewing Mr. Asay's paper. We are glad to see that someone who once took a clear anti-GPL stance now claims an "admiration" for it. We noticed the paper contained a number of email quotes from our founder, Richard Stallman, and our General Counsel, Eben Moglen. The quotes appear to be taken from various email conversations that Asay conducted with Moglen and Stallman separately. Quoting email out of context is always tricky, and the conflict and confusion with our official statements regarding the GNU GPL are apparent rather than real.

    We do, from time to time, have occasion to clarify our views on the GNU GPL, and we work hard to be responsive to questions from the community on them. This does serve to generate a lot of email traffic concerning the Foundation's opinions on GPL, and out of context pieces of the whole of such correspondence can be confusing. When we get pieces together that are particular good and clear, we post them on our philosophy page [gnu.org], as we did with Eben Moglen's essay, Enforcing the GNU GPL [gnu.org], and in our GNU GPL FAQ [gnu.org].

    As for the GNU GPL, version 3, we had indeed long ago seen Mr. Harris' proposal, but we did not feel it was representative of our plans for next version of the license. The copyright of the so-called draft of GPL Version 3 distributed alongside the article, thus, is incorrectly attributed to the Free Software Foundation. We do not hold copyright on it, nor did we have a hand in drafting it.

    Efforts for development of real GNU GPL, Version 3 continue within the Foundation. We of course expect to have extensive public discussion before any draft is finalized as the official new version of the license. In March 2002, We approved the Affero GPL [gnu.org], which includes a draft of one major provision we are considering for the next version of the GNU GPL. Of course, when we do announce that a draft copy of actual "GNU GPL, Version 3" is available, it will certainly come directly from the Foundation.

    Bradley M. Kuhn
    Executive Director, Free Software Foundation

  • Remarkably, after all this time, he's still only partially clued. As might be expected when somebody was totally wrong and announces the fact, when he explains his new understanding there's little more reason to take him seriously than before.

    From such little details as getting the names of the licenses wrong -- leaving the word GNU out of the title of his paper, and elaborating the LGPL as the "Library" GPL rather than its correct name, the Lesser GPL -- to totally missing the true competitive value of the GPL to businesses, as well as the very simple economic motivation for developers to participate, he still has a lot to learn.

    If there's anything new or interesting here, I didn't find it. He's articulate, though, and evidently open to reason after he has been pummeled sufficiently, so he might even get it right on the third try. In any case he's no worse than ESR.

  • That reputation argument is a crock. It sounds like some kind of monetary surrogate in the minds of people who can't organize reality in any other way.

    I participate in open source because I still believe in the tooth fairly. I put a tooth with a cavity under my pillow at night, I wake up the next morning and the cavity is gone.

    Once upon a time it was possible to have incredible wealth, yet still have bad teeth. You don't brush your teeth to improve your reputation. Perhaps those who bleach their teeth are in it for personal profit, but the rest of us just want good teeth _without_ spending a fortune on dental work. Is it really that hard to understand?
  • I completly lost the author when I read this quote:
    Very few applications exist in the open-source world, at least those that would be useful to the average user.
    Completely false! Especially for "average user". Tons of applications to suit all my need were installed automatically when I installed Mandrake. And I could easily complete by looking on the web (gimp, OpenOffice).

    The average user needs a browser. Konqueror, or Mozilla, or derivatives will do.
    The average user need a mail reader. Mozilla, Kmail, will do.
    The average user needs a word processor / spreadsheet. KOffice, or OpenOffice, will do.
    The average user needs a MP3 player, and instant messaging client, .... There is some on linux already.

    What else do you REALLY need? The author certainly didn't look much before making this affirmation. There are plenty of applications for Linux. And more are coming our way every day.

    The author also says that the GPL is an incetive killer for developper. For COMMERCIAL applications only. The way I see it, there is a lot more programming done with no commercial intentions. I like to code; and I always though of releasing some of my stuff as shareware or freeware. GPL is even better! Someone might improve it! What I think the author doesn't get is that the GPL intends to change the buisness model of software makers. And that is a good idea.

  • Holes in his proposed GPL... (Score:3, Interesting)

    by anthony_dipierro ( 543308 ) on Wednesday August 28, 2002 @06:12PM (#4159599) Journal
    • AS IS isn't capitalized. The UCC requires for it to be.
    • It doesn't define "source code".
    • It doesn't require a distributor of a derivative work to distribute that derivative work under the terms of the GPL unless that distributor is the same person who created the derivative work.
    • You only have to include a notice describing the changes you made, not the changes made by others. That effectively makes the requirement meaningless.
    • A distributor cannot impose license restrictions, but the creator of the derivative work can.
    • Clause 5 is equivalent to a restriction against removing the "do not remove" sticker on a mattress. I should be allowed to do whatever I want with my copy of the software.
    • I'm sure there are others.

  • GNU advocates stealing Apache credit (Score:3, Informative)

    by joneshenry ( 9497 ) on Wednesday August 28, 2002 @08:20PM (#4160332)
    Yet another GPL/GNU advocate steals credit from The Apache Software Foundation while claiming that Apache's success supports the GPL. Ironic how the same movement that is trying to advocate the name GNU Linux is shamelessly citing Apache's success as its own. Of course the Apache Software License is incompatible with the GPL so that code from the Apache and GNU projects cannot be mixed whatsoever, but that doesn't stop the GNU advocates from citing Apache as support for the GPL.

    How many more of these articles are we going to see that make bizarre claims like Apache is GPL software [newsforge.com]? (Note the passage containing "today thousands of GNU GPL software packages are already available including operating systems (like the famous Linux OS), office packages, Internet servers (like Apache running on 30% of Internet Servers".

Slashdot Top Deals

Beware of Programmers who carry screwdrivers. -- Leonard Brandwein

Close