Daily Archives: December 1, 2006

The TWIT license: Departure from the norm? What norm? Whose norm?

In a previous post I wrote about a new license I had created, the TWIT license. It consists of a small change to the MIT License. I changed the sentence

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

by deleting the phrase “and this permission notice” so the sentence reads:

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

I just deleted a few words. But just that act raises several questions.

Just what did I do? What does it mean? Did I have a right to do it? If not, whose right is it?

Simply put, I changed some writing. By deleting a few words I created a new form of that writing, what is called a “derivate work.”
I didn’t change my own writing. I changed someone else’s writing. And any form of writing — whether it be a novel, poem, play, piece of music, piece of code, the license for a piece of code, or a blog post about how to change the writing that is a license for a piece of code — has an owner, and the owner has the copyright of that writing.

And the copyright owner gets to decide which if any use others can make of the writing, and expresses that decision in the form of a license.

There are some exceptions. For example, you can quote an exceprt from a piece of writing as an example or to talk about the writing, under the notion of “fair use,” but you can’t copy the entire work without permission, as that would be “unfair use,” or at least that’s how I understand it.

What was the effect of my change?

In brief, both the MIT and TWIT licenses have several parts:

  • A copyright statement, “Copyright (c) …”
  • A list of things you can do, “Permission is hereby granted …”
  • A list of the conditions, or actions, you must do if you want to one of the things on the list;
  • A disclaimer, “THE SOFTWARE IS PROVIDED …”

My change only changed the list of conditions. I made the list shorter by dropping one of them.

I think it safe to say the TWIT license is more “liberal” than the MIT license. You can do everything under TWIT than you can under MIT, so I haven’t taken anything away. I’ve just reduced your obligations.

Now the MIT license is an “open-source” license because it has been declared to be one by the Open Source Initiative (OSI).

OSI is a non-profit that was set-up to promote and manage the notion of the term “open-source.” The OSI has created a definition of what it means to be “open-source,” called the Open Source Definition (OSD), , as well as an license approval process by which licenses may be reviewed to see if they meet the terms of the OSD, and a llist of OSI approved licenses. According to OSI a license is “open-source” if and only if the license is on the approved list.

In creating the TWIT license, I started with the MIT license and then made some changes. The MIT license is approved as “open-source.” Is the TWIT license also approved as “open-source.”

In making those changes I departed from the norm established by OSI.I started with MIT, an “open-source” license. Is the TWIT license “open-source.”

I started with the MIT license, an approved open-source license. I made some changes and produced the TWIT license. As I’ve noted the only effect of my change was to ask less of potential users, so I can argue that since the MIT license met the OSI definition, and since I made no change inconsistent with the definition, then the TWIT license also meets all the conditions of the OSI definition, and so the TWIT license is indeed “open-source.”

But can I say the TWIT license is “approved?”

If you and I can decide on our own what it means to be approved open-source, then why do we need OSI? We can sort all this out among ourselves, letting reason prevail.

Others make similar arguments. For example, robbers have as their business model the goal of taking your money. They do this by making an argument you will probably find quite reasonable and persuasive — that your life is worth more than the money they want. You can call their bluff and rely on reason to prevail … once.

So for “approval” to make any sense you need to have an “approver.” Indeed, the approval is worth no more than the reputation and authority of the approver.

Here again we see that open-source is not as simple as it first seems. (Though this is good news for people like myself who blog on open-source topics.)

I had thought of sending off an official requet to the OSI to see if I could get the TWIT license approved, and I’ll be writing more about OSI and approval later, but for now I’ve struck a middle course, by offering the content both under OSI-approved Apache2 the those who insist on an approved open-source license, and the TWIT license for those who are comfortable using that license.

We license your departure — If Nancy departed early, what gave here that license? What were the conclusions that Nancy drew?

In a prior post licensing issues that was inspired by a simple request from an editor named Nancy who sought permission to republish one of my blog posts, I observed that Nancy had probably not even read all of that post to realize that she had all the license she needed.

So here is the question. If Nancy didn’t read all the way to the end, then what were the conclusions that Nancy drew that led her to this decision?

It’s an issue that confuses many folks, so if Nancy got it right, I’m guessing she is a fan of the Nancy Drew mysteries, perhaps because the heroine is named Nancy.

So the question is now, what were the conclusions that Nancy drew, Nancy Drew being her inspiration?

If you go back to my reply to Nancy you will find the following phrase (emphasis added):

This should give you the permission you need. If it doesn’t, then you have it by this note.

That phrase gave Nancy unrestricted rights to use the blog post. I wrote the post, so I am the copyright owner, and the copyright owner — and only the copyright owner — gets to define the license for their work.

For example, if you own a piece of code, and then release it in open-source form, then doing so doesn’t limit your rights to the code.

This is a common practice. One term for it is “reference implementation.” You release the code but don’t accept any changes, and as long as only you make the changes yourself then you still have full ownership rights.

That is a key point. When you release code in open-source form the license doesn’t really kick in until you accept the first contribution. Then the license takes over. The only way to get around this is to acquire the rights to an earlier version, as was reported in the post What if? What not? Testing … testing … testing.

Another example. Most people probably believe that since MySQL is a widely-used open-source package, and that the company that has developed around it, MySQL.com, is based on that project, it is also an open-source company.

Howver, tt can also be argued that MySQL.com is a commercial software company. Just like Microsoft or IBM, it markets its code under a commercial license; that is its business. What most people think of as an open-source project is just the periodic publication of their code in reference form under an open-source license.

There is an easy way to test this hypothesis. If you dig under the covers you will find that MySQL.com employs all the developers. They have to, because they need to own the copyright to be able to offer that code under a commercial license.

Or if all the developers don’t work for them, then I expect that if you look into the way they accept contributions you will find that they request you sign over the copyright to your contributions. They might even say that are asking this as a favor to you, to save you “needless paperwork.” You will also find by digging into the past that they have purchased code for inclusion into their product. As I recall that’s how the acquired the code for the “innodb” engine. It is not common for open-source projects to buy code; much more commonly they solicit voluntary contributions.

There is an acid-test for this. Suppose write code that makes MySQL run 10 times faster in 10 times less memory. You wrote it all yourself. You have even published it on the web, so everyone knows about your magical patch.

You offer it to MySQL.com, but you refuse to assign them the copyright. If they accept it they won’t be able to incorporate it into their commercial code, because your patch would come to them under an open-source license.

So they can’t take it. So you simply make a copy of their last reference implementation, apply your patch, and distribute an offering that will cause real — perhaps even fatal — angst to MySQL.com. You may have a harder time making money, but at least you won’t have them around competing with you.

I’ve had prior experience in giving unrestricted rights. See for example the post Me Tube. It relates how I cause one web site to have unique authority in the display and use of the IBM logo.

And since the recent Sun / FSF alliance announcement of the release of Java in open-source form was not also accompanied by a statement that Sun would no longer be marketing a commercially-licensed version of Java, I bet that if you read through the agreements you will find that if you make a contribution to Sun/FSF then you will also have to sign over the copyright to your code, thus allowing the alliance to include your voluntary contribution in their commercial offering.

  • Pages

  • December 2006
    M T W T F S S
  • RSS The Wayward Word Press

  • Recent Comments

    daveshields on SPITBOL for OSX is now av…
    Russ Urquhart on SPITBOL for OSX is now av…
    Sahana’s Respo… on A brief history of Sahana by S…
    Sahana’s Respo… on A brief history of Sahana by S…
    James Murray on On being the maintainer, sole…
  • Archives

  • Blog Stats

  • Top Posts

  • Top Rated

  • Recent Posts

  • Archives

  • Top Rated