Open Source And The Obligation To Recycle

Lisa writes "Tim O'Reilly has a piece called "Open Source and the Obligation to Recycle" in his weblog, where he urges every company whose products are "obsolete" to consider making them available under an open source license, or putting them in the public domain, thereby enriching the soil of our collective commons. (Interestingly, the first posting on the weblog disagrees, saying "...Giving away the software of failed companies could turn every corporate failure into a disaster for everyone else.)""

Comments

Silly counter-argument

[kitts]

Giving away the software of failed companies could turn every corporate failure into a disaster for everyone else.

I'm probably taking this out of context, but this is a silly thing to say:

1. Corporate failures are not directly tied to bad software.

2. You can still learn something from the source code of bad software, even if it's only what not to do.

Re:Silly counter-argument

[Cato the Elder]

You are definetely taking the remark out of context...of course, so did the original poster.

I highly recommend you read the post, as it's fairly well written. At the risk of creating a strawman, BrettGlass argues that making failed software free will tend to hurt for-profit development in the same area, since they will still have to charge for their products. He asserts that it would be even worse if the orphaned code was relicensed under the GPL.

I have to disagree with him on the degree of harm that would be caused to commercial development efforts by this. The market has already shown a tendancy to go with commercially supported solutions. Now, as far as a brand-new product that just got dropped, yes, that could have a chilling effect on the market for a while. But that's not the kind of orphaned code the article talked about most. Besides, if the product could cause your companies death, don't you think you'd try to raise the money to buy the code?

I agree with him about releasing the code under the GPL, however. Doing that would prevent any commercial company from incorperating any of the code in there projects. At the same time, there would be no existing community of GPL-developers willing to work on the project and familiar with the codebase.

Re:Silly counter-argument

[dublin]

MS didn't give IE away they just made it part of the windows tax.

That's NOT what they said under oath in court: For those of you doing the Internet long enough to remember, you may recall that MS wasn't up to writing a browser of their own to challenge Netscape: So instead, they decided to buy one (or actually, steal one, as you'll see in a moment.)

The only thing that was even close to being a Netscape competitor in those days was the original NCSA Mosaic code, which was spun off for commercialization by UIUC(.edu) as a company called Spyglass. Spyglass tried unsuccessfully for a while to land big buyers in hopes of competing with Netscape, but their code wasn't nearly so good as the Mozilla crowds' (back before Mozillla meant open source...) Finally, they landed the biggest fish of them all, Microsoft: They struck a deal with Microsoft to be the Microsoft browser: with backing and volume like that, they couldn't lose! Spyglass poured millions into develpoment and features that Microsoft wanted in the product - they knew they'd get their money back because the contract with Microsoft guaranteed them a percentage cut of every copy sold.

But Microsoft NEVER SOLD A BROWSER! Instead, it simply became "part of the operating system" (avoiding having to pay Spyglass was one of the biggest reasons BillG wanted to claim this.)

There was, of course, a law suit about this, which Microsoft won by swearing that since IE was an integral part of the OS, and not something that was even possible to buy separately, they owed Spyglass nothing for the millions of copies of their code that they distributed: Since they had'nt sold any IE they owed no royalties! Microsoft won leaving Spyglass with nothing for all its hard work and destitute to the point that they finally had to sell out to OpenTV in the hopes of becoming a niche browser for set-top boxes...

Problems with this

[Com2Kid]

I remember reading up the other day on a game called Star Command.

When asked why they didn't release the game for free or open source it, a person in the company (CEO/Lead developer or some such high up person, small company anyways) said that it would take too much of their time and money.

First they would have to try and FIND the source code (doubtful if it still existed), then if they managed to do that, reassign some people from another money making project to taking an old (in this case DOS game) and removing the protections and security checks they put in there and scanning in the documentation, bundeling it up, setting up a server for distribution (or maintaining a sourceforge account, granted they could pass it on to somebody, but then that entales the legal issues involved), all just to make a few people happy that they could now freely get an old DOS game.

Sure it /MIGHT/ be upgraded to something modern and ported all over heck and such, but shoot, reality is that it would cost them MONEY when they are already a struggeling small company.

Not to mention cases where the rights to a program are split across a few gazzilion people and numerious corporations. This is especialy true when one company has the rights to the game and another company has the rights to distribution. Icky situation there. And if by some chance somebody sold off merchandising rights. . . . oh man, no hope at /ALL/.

A good first step though would be to /REDUCE/ the copyright limit. ~7years for computer programs and ~25 years for books and other documents sounds nice. (Some books stay in print and keep on selling for longer then that though)

At least it would take care of the legal hassles somewhat, but it still wouldn't help with finding the sourcecode.

"umm, lets see now, where did I shove that 5.25' disk. . . . . "

Reply to BrettGlass

[ryants]

Ironically, the failure of my competitor is likely to kill my company, too -- EVEN IF MY PRODUCT IS SUPERIOR. (People will tolerate many shortcomings in something that's free.)

Then why doesn't Linux (*BSD, etc) own 95% of the desktop market?

The situation gets worse still if the GPL enters the picture. If the competitor's code is released under the GPL, I cannot so much as LOOK at it.

Sure you can. Just don't copy it.

Thus, the sudden release of software from companies that go out of business either into the public domain or (worse) under the GPL can cause a chain reaction which destroys any incentives to create or improve products in that category.

*laugh* Really... Microsoft continues to "compete" with Linux, KOffice, etc. Eudora competes with mutt. I'm afraid I just don't see any justification for the above statement.

Re:Why dosen't he follow his own advice

[checkitout]

But if you look at the complete Open Books list, you'll also see a number of out of print books. These books were open sourced not because we wanted to spread the software or the ideas, but because we felt an obligation to make the material available to those who could make use of it even after we were no longer able to sell the books profitably ourselves. This is recycling in action.

If had you bothered to read the article, he mentions that he has.

giving s/w away is not a disaster for others

[markj02]

Giving away the software of failed companies could turn every corporate failure into a disaster for everyone else.

I presume what the poster means is that by giving away the software, the company destroys the market for the competitors.

Well, that isn't quite true, as we have seen again and again. In fact, in real life, the source code and executable are only a small part of the value of a software product. Most of the value is in the ongoing maintenance, business relationships, trademarks, the user base, the books that have been written about it, in short, the "network" that surrounds it.

To the degree that it is true, well, software companies simply have to get used to the fact that, once created, it costs nothing to give software to additional people--that ultimately has a lot of influence over how software can be priced and licensed. There is no use whining about basic economic realities.

Failure of copyright to establish a public domain

[jms]

This article illustrates the complete failure of the only constitutional purpose of modern copyright law with respect to software -- a failure to establish a public domain, in both senses of the word.

The first sense is the idea of public domain as "uncopyrighted" or "expired copyright." Had Congress resisted the urge to tamper with the copyright laws in 1976, things would be different. Under the pre-1976 copyright regime, copyright lasted for 28 years, with the option to re-register for an additional 28 year term. Under this system, abandonware from the early 1970s would be now regularly entering the public domain. In two or so years, we would start to see the first generation of abandoned PC software enter the public domain -- old Apple II software, games and system software from long-lost companies. Instead, by repeatedly extending copyright, and removing the renewal requirement, Congress has essentially consigned the history of computer software to destruction. Very few of us will live long enough to be allowed to legally copy the software that was written before many of us were even born. Even if we did live long enough, the media will have long decayed, any software from the early days of personal computers will only survive as illegally made copies. In essence, Congress has criminalized the work of the historian and archivist, with no real benefit to anyone.

There is another sense of the "public domain" in which the copyright laws have even more drastically failed. This is the sense of the "public domain" as "the body of work available to the public to read and learn from." The problem is that by allowing copyright on object code, and by allowing software publishers to treat source code as trade secrets, in essence, computer programmers are forced to learn their trade from scratch. Imagine if a student expressed an interest in becoming a writer, and was told, "If you want to be a writer, you will never be permitted to legally read books written by successful, popular authors." I doubt that the result would be better literature, but that our public policy with respect to software -- both by attaching copyright protection to object code, and by allowing the attachment of licenses to software that forbid reverse-engineering -- a technical term for "reading" software.

The primary purpose of copyright was to place knowledge into the public domain. That's why patents must be openly published, and why, originally, only published works were eligible for copyright. Now, with copyright protection automatically attaching to all works, whether or not they are ever published in a form that adds to the public domain, we are back to the bad old days of proprietary licensing of -- and the subsequent destruction of learning and knowledge -- the very problem that copyright was designed to put an end to!

I believe that the real revolution in free software is not a better business model. It is not the sense of community that it fosters. It is not the reduced costs or the improved quality.

The real revolution in free software is that it in effect reestablishes the public domain that has been systematically destroyed by Congress in passing ever more restrictive, destructive copyright legislation. In the year 2002, free software is the public domain. It's the software that you can download, study, modify, improve, sell, and give away. It's the software that you can learn from, instead of just use.

Unlike proprietary software, free software is the software that promotes the progress of science and the useful arts, and anyone who is interested in promoting progress in the field of computer science should strongly consider releasing their software under the GPL, after its commercial potential has been exhausted.

always costs money to do this

[egomaniac]

The request sounds simple in theory, but as always it's more complicated than this.

What if somebody is scanning through the code and finds

// we have to do it this way because Intel is a bunch of fucking idiots

Would you really want that being leaked into the public? Particularly if you have a close working relationship with Intel?

Or how about:

// Bob Martel isn't at the company anymore, so I can safely say that this is the worse piece of shit code I've ever seen. I'll rewrite it when I get a chance

Not to mention potentially plagiarized code, or patent violations, or any number of other nastiness buried in the code. Sure, it *probably* doesn't contain any of the above, but do you really want to scan through all two million lines of it just to make sure?

Further, every time somebody wants to see my source code, I find myself preparing it a little more thoroughly. Removing a few of those ugly hacks, documenting the ones I can't remove ... sort of like quickly straightening up the house before visitors show up. I don't want people seeing my dirty laundry. If I don't have time to straighten up, I'd rather not let people in at all.

Source code is the same way -- you generally don't want other people looking over it until you've had a chance to clean it up a little bit. If you don't want to clean it up, you just don't release it. Releasing source code *always* costs time and money to a corporation.