It’s free and open for a reason!

Just as I was writing this post on ignorance of open source models, there was news of a major publisher making a print edition of Wikipedia. This led to backlash reactions which illustrate exactly the kind of misunderstanding that arises when people don’t get open-source licenses (via ReadWriteWeb.

The whole point of Wikipedia is to provide a voluntary online source of information. That is why people voluntarily gave up some of their time to write the articles for free in the first place. It is a noble project, the writer believed in the project and so they participated in it. Had they known that Wikipedia would then use their work in a commercial printed venture, I’m sure they would have had second thoughts about writing those articles. At the very least they would have demanded a contract and perhaps a guarantee of financial compensation later.

Now the author might well have a point – maybe people don’t know what they’re doing when they’re adding to wikipedia and maybe wikipedia should make it more clear. Still, I find it worrying that they would consider some big ball of wrong has been cast upon them. A great place for them to start researching would be the GNU article on wikipedia, the one article whose authors knew they would be victims right from the start ;).

There are several disturbing things about this:

  • First, if this ignorance does exist, that’s unfortunate. If anyone is a major enough contributor to wikipedia to care about this, they really ought to familiarise themselves with the basics of the GNU license they’re operating under. (And for the record, I think it’s a shame wikipedia uses GNU instead of a more permissive license.)
  • It’s scarcity mentality. Boo-hoo, the publisher is making money and I’m not, so they must be ripping me off. Yeah right! If you were willing to contribute to wikipedia before, your content remains just as available online as before. What’s changed? Nothing, except someone else is going to make your content more digestible in certain contexts and accessible to more people. And yes, they will make money from it. Maybe they’ll make lots of money. Whatever. That’s what happens in a modern knowledge economy – people like Larry and Sergey make money without denying it from someone else. Far from removing others’ assets, they actually add value to everyone else. Likewise, the existence of a print wikipedia will also add, in its own way, to society’s wealth. (Although not the forests. That’s a separate, more legitimate, argument.)
  • There’s an old psychology result that says people care more about their spending power in relative terms than absolute terms. ie you ask them “Would you rather earn $50,000 a year while other people make $25,000, or would you rather earn $100,000 a year while other people get $250,000? Assume price of service and goods would remain the same.” People would rather earn half as much, if it means they are richer rather than poorer than others. That result says a lot for this kind of irrational backlash against anyone who might cash in on their “volunteer” work.
  • It’s hardly the first time this has happened. and hundreds of fairly useless SEO sites have been profiting from wikipedia content for years. I’m surprised it took this long for someone to do it. I could imagine countless opportunities for books on niche content – bizarre trivia, country almanacs, movie guides, etc etc.

That a company can do something like this without having to formally negotiate with the wikipedia foundation and a million contributors…I call that a cause for celebration, not outrage. Personally, I’ve contributed various boring tidbits to wikipedia and I hereby invite any publisher to print what they want from my contributions, even if there’s only limited demand for hard copy about snowclones.

Models of Open Source

Controversy arose in several open source projects last week: Ext JS, SpringSource App Platform, and the SWF/FLV specs. Dion has a good summary and concludes with a reminder that most people could care less:

Finally, I know that 99% of the developers out there may not even care, let alone users. There are open source wonks who like to argue about licensing and methodologies. As I watched the John Adams HBO series, I felt a little like those fine chaps arguing over the finer details of things. Many of the people didn’t know what was going on there, or why a particular Article was written the way it was. But, they had drastic implications for the people. I think that the same can hold here for some of the projects.

Now this ignorance of open source models in my experience and an interesting phenomenon. It took open source until around the middle of this decade for even the basic definition of open source to be understood, and to some extent, accepted, by mainstream IT and business communities. That’s about five decades after the practice began, two decades after Stallman’s GNU Manifest, and almost a decade after Raymond et al declared “Open Source” teh nu k00lness.

The basic definition of open source is understood, but as Dion notes, most people have no idea how the models vary, or what the implications are. Which is unfortunate, because the commercial implications are indeed vast.

I consider myself an expert on this topic ;P because I once picked up a copy of an O’Reilly book on open source models at a heavily reduced price and flicked through it intensely for at least two hours. Okay, so not quite an expert (!), more an armchair ignoramus, but certainly a stakeholder as both a consumer and publisher of open-source. And I’m hardly the Richard Stallman of the Ajax world either, but whenever I publish software that could potentially be used or adapted by someone, I make the tiny effort of including a license notice. Sometimes it almost seems arrogant to do so when the software is so trivial, but I’d rather do that than leave the license unstated, in which case it would come under the silly default, i.e. standard restrictive copyright. (Even if you agree with standard copyright, doesn’t it seem silly that rights to anything anyone creates belongs to them by default, and for the greater part of a century?)

Well, since I include a notice, I’ve had to think about which license to use.

In my dim-witted view of models, there are three key categories. All allow people to use it free of charge and generally remove any liability from the provider. They vary in what happens when people change it:

  • Copyleft/Viral (e.g. GPL, Creative Commons Share-Alike) – Anyone who modifies it must release the modified version.
  • Permissive (e.g. MIT, BSD, Apache) – You don’t have to release your modifications, though you still have some restrictions such as retaining the original license and author details.
  • Public Domain – Anything goes.

There are other distinctions too, e.g. around charging and attribution to the original creation; but the above distinction is the most important in most cases. And this is the sad thing: most IT and business people simply don’t understand these distinctions and the implications they have.

You can argue about ethics until you’re blue in the face, but the reality is, there’s no absolutely “right” or “wrong” model. It’s just horses for courses and being aware of the implications. Personally, I do have a hard time justifying copyleft-style licenses and I sometimes wonder how many people release stuff as GNU just because they associate it with open-source, without realising the full implications. I’ve worked in many commercial situations where GPL is unacceptable and I hate the idea that developers will be in the same situation as me, i.e. gleefully eyeing something that would scratch their itch, but not being able to use it. Incidentally, Ajax Patterns was initially released with the more permissive Creative Commons license, but following the book deal, O’Reilly asked me to change it to share-alike as it discourages other publishers from publishing their own version of the same book, which has apparently happened in the past. This is similar to the reason some companies release dual licenses – GPL and standard copyright. It means anyone can use it for free under GPL, but other companies can’t update and re-distribute for a fee. I can see that. It creates an incentive to release as open-source.

It’s convenient for me to have a default license choice, so I don’t have to think too hard whenever I release something. I have chosen The MIT License because it is extremely concise, and is basically the smallest (in words) mainstream license that satisfies my main aims: To allow people to work with it freely and free of the restrictive burden GNU creates; to receive credit for my work; to be free of liability. The three paragraph license can easily be embedded in any source file, which is great for releasing one-file libraries without having to include a separate README or LICENSE file. Best of all, it’s easy to grok. If developers should be able to consume open source as easily as pulling out books from a shelf, you really don’t want to release software under a license that only a lawyer could love. Don’t let MIT’s concise nature fool you into thinking it’s only used for toy projects: MIT is used in Rails, Mono, and the system for which it was originally written: X-Windows.

The MIT License looks like this:

Copyright (c)

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the “Software”), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.