Thoughts on "open" or "free" licenses for documentation

Posted 12 Aug 2001 at 02:02 UTC by Rich Share This

The Reefknot project ( is currently discussing what license to slap onto the a certain part of the documentation itself. Licenses being considered are the Gnu Free Documentation License, and the Open Publication License. I wrote the following to the mailing list, and got a little carried away, so thought it should have wider reading. This article is released under the Open Publication License

On Sat, 11 Aug 2001, srl wrote:

> After some discussion on IRC, I've added the Open Publications > Licence to CVS, and set things up so the Bootstrap Guide is licensed > under that. It seems clearer to me, and it's still Free by the GNU > definition so long as we elect out of the options. see > doc/bootstrap-guide/LICENSE for the details.

After further discussion on IRC, I thought I should weigh in with my opinion here. I'm sorry if this becomes a treatise. For those with little patience, I've placed an executive summary at the top.


I like the OPL better than I like the FDL. It is more clearly stated, shorter, a lot easier to understand, and seems less pedantic about the specific details of what is and is not allowed. I.e. it seems more free, ironically.


I come at this whole license discussion from three angles, which I will attempt to correlate with three of the seven deadly sins before I'm done. As an avid fan of Charles Dickens, himself a staunch supporter of Copyright Law (of which there was none by his death in 1870), I will talk about GREED, aka COVETOUSNESS. As an author myself, I will talk about HUBRIS, aka PRIDE. And as an author of a derivative work, I will talk about either ENVY or SLOTH.

Now, although this is shaping up as a sermon, take heart, I think I'm going to end up saying that these things are all really good things, and should be protected by our licenses. And I'll try to squeeze ANGER and LUST in there somehow if I can.

Please stick with me. I will be making some points.

Dickens: (GREED)

Charles Dickens wrote a few books. Most literate people have heard of one or two of them, and many have even read one. I have read almost all of his books. And believe me, there were a crapload of them.

In 1850something, Dickens came to the United States of America, where he was treated like a celebrity until he told folks what he was here about. He was concerned that whenever his books arrived in the USA with his official publishers, everyone had already read bootleg copies of his books, from which he made no money. He was an advocate of Copyright Law because he knew that such a law would ensure that he would be rediculously wealthy.

Many content licenses protect the ability of the author of the content to make a living off of the content. This, in turn, encourages authors to write good stuff. Dickens argued that open content licenses - the ability to reprint an author's work without 1) his consent, and 2) without paying him anything, will stifle innovation. (Any of this sounding familiar?) That is, authors would have no incentive to write good stuff.

His detractors, by they way, said that anything written purely for money was crass, and only "ars gratia artis", or art for the sake of art, was ever going to be really good art. His detractors, not coincidentally, were independantly wealthy.

The point of all of this is that I think that a content license should protect the ability of the authors to make money off of that content if they choose to do so. This may be labelled as GREED, but it is what enables us to live comfortably, take care of our families, and buy nicer computers.

Both the FDL and the OPL (or OCL) protect this right, as I understand them.


When I write something, I want it to be appreciated. In fact, when I write, 90% of the reason for doing so is that I want to do something that people notice - something that is important, makes an impact, and helps people learn something useful. This inevitably frustrates my publishers, who want to get a book printed by the end of the fiscal quarter.

A content license should protect my pride, in the following way. It should be difficult to do two things.

First, it should be difficult for someone else to take my work and republish it without attributing to me the glory due me. That is, if they reprint it, and say that they wrote it, then they are dealing a blow to my pride. This is called plagiarism, and is just bad form.

Secondly, it should be difficult for someone to take my work, screw it up, and then make me look stupid by releasing the mangled version and attributing it to me. I embarrass myself quite well enough without your help, thank you very much.

As I understand them, the FDL and the OPL protect me from this, but in different ways. The OPL requires that if someone makes changes, they tell me about those changes, that they clearly state that their version is modified, and that they clearly indicate where the original can be found. And, it requires that if they publish it that they let me know before they do so, in order that I can make sure that what they are publishing accurately reflects what I wanted to communicate. This is a Good Thing, and adequately protects my PRIDE. The FDL is much less clear on this point, and unless someone can explain it to me otherwise, seems only to say that they are required to state where the original is, and only required to tell me if they distribute more than 100 copies. And it's not clear to me that this is required, just recommended. On this point, I think that the OPL clearly has the advantage.

Derivative works: (ENVY, SLOTH, perhaps others)

I am the author of Apache Server Unleashed. This is basically a 800-page book in which I took the Apache Server Documentation, and made it more palatable to the common beginner Apache admin. I did end up contributing some of this back to the Apache docs, and will probably do even more of that as I work on my second book, but I'm not sure of that point adds to, or detracts from, my argument.

The point that I'd like to make here is that the license on the Apache documentation, which, as nearly as I can tell, is the same as that on the Apache code itself, permitted me to get away with this. The license on the Perl documentation, and on the various docs on CPAN modules, on the other hand, would take a team of lawyers to figure out. And this is stated in the CPAN docs somewhere.

While I ENVY the Apache authors their technical skills, I was able to excercise my SLOTH to produce a book based on their work - primarily the work of Ken Coar, who was gracious enough to be my co-author on that book - and was able to make a tidy sum of money off of doing so.

A content license should protect the ability of the *unrelated* author to produce a book based on that material, and to sell it, and to make money on it, and to not give a nickel back to the original authors of the document.

While it sounds, initially, that this is directly at odds with my first point, I don't think that it is, because an author with any integrity whatsoever will not simply republish that document, or cut-and-paste large portions of that document into their own work, since that is plagiarism. An author with any integrity will also feel obliged to contribute back to the original project, although they should not be required to do so. They should, however, be require to contact the original authors and tell them what they are doing.

Perhaps this is at odds with my first point. I'm not entirely sure. This needs more thought.

However, back to the matter that we are debating. The FDL, it would seem, almost expressly prohibits this sort of thing, while the OPL permits it, given that the author of the derivative work notifies the original authors, and submits any modifications that come out of the process.

And, of course, the original authors should have the foresight to publish a book of their own, and beat everyone to the punch.


I believe that the Open Content Licsense, or whatever it is called these days, more accurately reflects my views on the topic of documentation licensing, and that the PDL explicitly takes away some of the things that I think are important, in addition to stating everything in terms that make my head hurt while trying to understand them, and wondering why they don't just say what they mean.

And, um, I did not manage to squeeze ANGER and LUST in there. For which I am deeply remorseful. Which is the appropriate reaction to sin.

Use the BSD doc. license, posted 12 Aug 2001 at 09:47 UTC by nik » (Master)

Like Apache, the license for the FreeBSD documentation is very similar to that of the code. It can be seen at

This does not preclude third parties from carrying out commercial printing of the documentation (for example, WindRiver Systems is preparing the second printed edition of the FreeBSD Handbook, and they are investing substantial resources in doing so), but it does mean that your copyright message must be retained.

Community vs Code, posted 12 Aug 2001 at 10:24 UTC by ask » (Master)

I've no idea how it applies to documentation licensing, but anyway...

One summary of the difference between how the GPL and the Apache Software License works is:

GPL is "viral" on the code.

Apache Software License is "viral" on the community (you can do whatever, except calling the new software for Apache and you have to advertise the existense of the Apache Software Foundation).

I believe that it's Roy Fielding who sometimes argues like this[1]: most open source projects can be clean room implemented in N months (with fairly low values of N) by a dedicated team of talented people, so following that the code is really not worth all that much. What's valuable is the community, so protecting the community and keeping the community working and getting "new blood" to it, is the important thing for open source projects.

- ask

[1] but I might very well be putting words in his mouth, so don't blame him if my words are not even close to what he says. :-)

Dickens, copyright, etc., posted 13 Aug 2001 at 18:49 UTC by jbuck » (Master)

OK, this is a bit off topic, but it addresses a point in the original article.

No, Americans weren't reading bootleg copies of Dickens. Rather, the US did not recognize non-US copyrights or patents until almost the end of the 19th century. Dickens was not legally entitled to one dime from his American readers or American publishers.

The US created its industrial revolution on "stolen" British patents, and US industrialists got their start by smuggling machinery out of Britain (since patents had no force, the British tried to protect their technical lead by forbidding exports of some types of equipment, and in some circumstances engineers and other technical experts weren't allowed to leave the country).

If the US had been forced to comply with current international law (the GATT and other treaties and trade agreements), the country would have remained backward (ok, I hear sniggering from our British friends) and agricultural for far longer, a colony forced to buy its technology from Europe. In short, the US got to its current competitive position in the world by means of what is now called piracy.

not piracy!, posted 15 Aug 2001 at 02:26 UTC by yakk » (Master)

Its only piracy if someone else does it, right?

Anyway, piracy is such a bad word for the breach of copyright law. Theres no murder, theres no rape, theres no stealing of money, gold or other physical wealth. All there is is perhaps a diminished possibility of sales and profit.

Check out CDL also for reference, posted 20 Aug 2001 at 21:25 UTC by wsanchez » (Master)

We went through some similar discussions at Apple about what license to use for Darwin doc. I think the upshot was that we like the GNU PDL except that it was kind of complicated and we just wanted something simpler which basically preserves credit. What we came up with is the Common Documentation License. Here's the preamble:

The Common Documentation License (CDL) provides a very simple and consistent license that allows relatively unrestricted use and redistribution of documents while still maintaining the author's credit and intent. To preserve sim plicity, the License does not specify in detail how (e.g. font size) or where (e.g. title page, etc.) the author should be credited. To preserve consistency, changes to the CDL are not allowed and all derivatives of CDL documents are required to remain un der the CDL. Together, these constraints enable third parties to easily and safely reuse CDL documents, making the CDL ideal for authors who desire a wide distribution of their work. However, this means the CDL does not allow authors to restrict precisely how their work is used or represented, making it inappropriate for those desiring more finely-grained control.

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!

Share this page