I guess it’s time to chime in one of the most heated Open Source debates, as I consider attribution a disaster waiting to happen: every potential scenario is probably going to split the Open Source community, raise a lot of concerns and bad feelings, and undermine credibility of the OSI. I really wouldn’t want to be in OSI’s shoes: these guys have their backs against the wall, as one one hand I don’t see any sound reason to reject attribution based on the Open Source Definition, and on the other I can clearly see how the Open Source community at large will react should attribution licenses be blessed and considered kosher.
Let me get away with just a few words on the practical matter: I consider attribution cumbersome, confusing, difficult to understand and comply to, and overall dangerous if you consider what it might bring to. A lot of people have been more vocal than me on the subject, so I won’t wast any more electrons on this. Then again, as much as I might consider attribution licenses a pile of fetid dingo kidneys, I see no strong formal reason to reject them from an OSD standpoint. And this is the real problem, as it clearly shows we have a problem with the definition itself.
The heated debate on attribution is clearly splitting the community in two, somewhat acting as the kid who shouted about the Emperor’s new clothes. Attribution fans mostly belong to that side of “business/professional” Open Source, typically VC-backed, who feels the need for protection, cares about the “quid pro quo” factor, and is scared to death of the Bogeyman running away with their souls and their source code. They are telling you they are just looking for credit where credit is due, they want to protect their assets from competitors and that attribution in the form of badgeware is really no different from attribution the way it’s stated in the BSD license, so they are entitled to it. They are not telling you how they are, in the end, old time commercial software makers in disguise, adopting Open Source since it is the only way to enter the product market nowadays, and how attribution fits just fine in their “let’s cripple our software so that users will upgrade to the commercial edition” strategy.
Opponents of attribution are mostly Open Source die-hards, who feel Open Source is a reward in itself, the “quid pro quo” is implicit in collaboration and virtuous cycles, and the risk of forks is something they can live and cope with given that good communities rarely split and forks might even be healthy sometimes. They are telling you they don’t like attribution as they are after collaboration, transparency and consistency to Open Source users. They are not telling you they want to use every bit of Open Source the way they are used to, possibly doing business with it with no strings attached, and they have a problem with commercial vendors who try to get in the Open Source spotlight, grab a free ride on the Open Source train, forget about participation, throw source over the wall and fly business class with their VC money while they strive on their shoestring budget.
I understand the description above is somewhat biased, but you know where I stand, don’t you? Let me however counterbalance that statement: even though I think those guys are working their way through loopholes of the Open Source Definition, I can’t really blame them for that. Actually we happen to work with at least one of them (with more to come), and I’m happy with the relationship so far. I do think they would be even more profitable and efficient if they’d take the red pill once and for all, but it’s their take and I respect their decision, no matter how I might consider their behaviour suboptimal and unfit to the Open Source the way I feel it should be.
Back to the split, though: I might sound like a broken record, but I just can’t miss such an occasion to remind my fellow readers how I anticipated these frictions a while ago. From the “Hah! Told you so” department, it’s good to see how Matt Asay himself is now both sharing my concerns about Open Source dilution and playing the “my Open Source is bigger than yours” game. As the Open Source landscape grows, and as the Open Source Definition gets more and more unable to describe in a coherent and specific way the different players, their motivations and their objectives, this kind of bad feelings is assured to grow, together with more and more confusion.
What then about users, confronted with different incarnations of Open Source that hardly have anything in common? What we are seeing now is far from beneficial, yet there is nothing we can do at this point other than help our users by working towards new taxonomies that can describe and accommodate the different visions, possibly keeping a common Open Source umbrella but also understanding how it might be doomed to be way too broad to be of any use in the mid-long term. Hey, Greenlandic has no less than 49 words for “snow”, so I guess there must be quite a lot of room to define Open Source better than it is now.
By the way, in case I wasn’t clear enough: attribution sucks.