Open Sources | Rodrigues & Urlocker » Attribution: Mis-pointing fingers

January 19, 2007 | Comments: (0)

Attribution: Mis-pointing fingers

TAGS: Open Source

Ca suffit! I have hit my limit on the attribution debate. I'm a little weary of people haranguing companies for using attribution licenses, primarily because it is the OSI, not the companies, that should be under pressure right now. A generic attribution license has been submitted to the OSI. It's been sitting there for months now, going through endless theorizing on license-discuss, with no concrete decisions coming out of the OSI. (Note: I'm an OSI board member and also work for a company that uses attribution, as such, I have recused myself from voting on the issue within OSI.)

So, if you don't like attribution, then pressure OSI to make a decision. The companies legitimately feel they are 100% open source. Some disagree. It's for OSI to decide at this point - not you, and not me.

As I've written, I'm no friend of attribution. I think it's a weak method to achieve aims that the GPL is better constructed to do. There isn't a day that goes by that I don't advise a company (including my own) to adopt the GPL.

But this doesn't mean that attribution isn't open source. There are plenty of open source licenses that I don't like, don't use, and would be happy to see go away. But they're still open source licenses.

Here's why I'm not buying the anti-attribution movement's open source argument:

• Attribution does not burden distribution (any more than many other open source licenses, including the GPL). Some of the criticism has claimed that attribution prevents redistribution. This is quantitatively not true. There are tens of thousands (perhaps more) of sites that use attribution-"burdened" software, many of which include composites of different open source projects. Whatever theory might predict, reality definitively shows that attribution has not unduly burdened redistribution.

• Attribution will consume the UI as different attribution licenses are combined. I grok this point (though it's more theory than reality). It is, however, not really any different from the existing fact that some open source licenses simply aren't compatible with others. (Having said that, I'll state for the record that I don't believe that Apache licensing truly is incompatible with the GPL, but there is a strong perception that this is the case, which prevents a lot of interweaving of the two licenses that might otherwise happen.) Perhaps you won't want to integrate Zimbra with SugarCRM with SocialText...and? This is not an open source problem. It's a practical problem.

• Attribution doesn't work in headless environments. This is actually a good point (i.e., can you use the software if you're using it headless, or in an environment where you don't have a UI, as it were?), and goes to #10 of the Open Source Definition. Perhaps it would require modification of these MPL+attribution licenses, but I'm not sure the baby should go out with the bath water.

• Attribution is consistent with OSD #4 (Integrity of an author's work), droit morale, etc.. While some in the development community don't buy this argument, I can tell you from actually sitting in on these discussions that it is a big deal for a company to see its software poorly supported. The OSI provides this rationale for OSD #4:

  Encouraging lots of improvement is a good thing, but users have a right to know who is responsible for the software they are using. Authors and maintainers have reciprocal right to know what they're being asked to support and protect their reputations.

  Accordingly, an open-source license must guarantee that source be readily available, but may require that it be distributed as pristine base sources plus patches. In this way, "unofficial" changes can be made available but readily distinguished from the base source.

Rather than spend our time counting how many attribution licenses would fit on the head of a pin, we should be focusing on the practical effects of attribution. I can understand that some developers prefer not to use software that requires attribution. I totally get it. But this is no different from me not wanting to use BSD-style licenses, Apache not wanting to use GPL-style licenses, etc. This isn't about preferences. It's about what is open source, and what is not.

I am a pragmatist, in the William James slant on the word. I believe in what works. If it works, it is right. Attribution has blessed us with mountains of exceptional code. Code from which tens of thousands of groups are deriving daily, tangible value (with similar numbers of developers involved in modifying, forking, and promoting). I know, for example, 6,000 active users of attribution-"burdened" software at Alfresco alone. Add that to the Zimbra, SugarCRM, SocialText, MuleSource, Intalio, etc. etc. users and you quickly get to a massive population of users and developers that don't seem to share the same distaste for attribution.

I guess this makes me an unwilling apologist for open source attribution. It's not something I prefer to do with my time - I don't like attribution, but also don't think that personal preferences should get in the way of pragmatism. Pragmatism in the Jamesian sense.

/Matt talking about attribution (hopefully forever)

Posted by Matt Asay on January 19, 2007 07:12 AM

RATE THIS ARTICLE: