Further Examining Microsoft’s CodePlex Foundation

The analysis of Microsoft’s newly announced CodePlex Foundation has started. I commented on this story in a recent link post.

First I found a posting by Josh Berkus at IT Toolbox via Hacker News. He points out some vague language around commercial software that suggests the true focus of the new foundation may at best be supporting interoperability between closed and open systems. The foundation itself is constituted as a trade association, not a traditional not for profit like comparable open source bodies. That reinforces his theory that the organization is about commercial interests, not open source principles.

I don’t have the same qualms about the contributor agreement. A non-exclusive license is not really all that unique but for a simple project hosting site, it is a little odd. Usually you sign onto such an agreement when contributing code into a specific project, like GNU or Apache, not for standing up your code on a site like SourceForge or github. I also found it odd that the agreement doesn’t speak to license compatibility of any kind. Berkus does get it right that Microsoft is granting themselves the ability to re-license your work under whatever terms they like.

The other analysis I saw was via a tweet from Glyn Moody. The link he shared pointed to a post by Andy Updegrove on a site, ConsortiumInfo, apparently much more concerned with constitutional and operational issues around industry consortia. Updegrove explains he looked at the foundational documents based on his 22 years experience participating in and running similar foundations, most notably the Linux Foundation.

He also spots the choice of 501(c)(6), a trade association, but seems far more sanguine with it. His comments are extensive and deep but in a nutshell, his qualms have more to do with how the board is formed, its decision making capabilities, and its relationship to general membership. He seems to think the problems he spots are fixable.

He concurs with my reading of the contributors agreement, noting that it is unusual to frame this as if incoming code is going into some general code bank, rather than a specific project. He also spots a few contradictory statements around the licensing stance and worse the fact that changing this language could be problematic, even at such an early stage.

As for the focus on commercial outfits, Updegrove also sees this a bit differently than Berkus. He suggests the exercise of replacing “commercial companies” with “Microsoft” and you’ll see pretty clearly their intention. Otherwise, he argues that most other commercial software vendors these days have more mature and thoughtful policies around open source contribution and collaboration.

Updegrove ultimately gives Microsoft the benefit of the doubt, basically citing Hanlon’s Razor: “Never attribute to malice that which can be adequately explained by stupidity”. In this case stupidity would be in contracting to some outside firm to put together the governing bylaws for this new foundation. He also provides a set of what he sees as critical changes to make the operation of the new body mesh with Microsoft’s stated high level goal.

I’ve said it before and I will re-iterate. Microsoft has a deep and ingrained antagonism towards open source in general and Linux specifically. This is not a feeling or an intuition but an observation based on a pattern of facts. Granted, they’ve at times tried to sow doubt around those facts but you don’t accidentally repeat these same sorts of what I’ll charitably allow as mistakes over and over.

I’d welcome a genuine change of heart on the part of the Redmond giant, really I would. I try not to have unrealistic expectations of the open source policies of companies charged to turn a profit. I think we’d all benefit by a minimum level of comprehension of the benefits of collaboration and sharing, along side traditional competition based models. Really I am willing to extend Microsoft that benefit of that doubt but to do so is going to take more than a clumsily formed foundation, a few vague promises, and some code contributions made without any effort to follow up and support them.

Leave a Reply

Your email address will not be published. Required fields are marked *