I’m very mildly annoyed with some discussion of “open data”, in part where it is an amorphous thing for which expectations must be managed, value found and sustainable business models, perhaps marketplaces, invented, all with an abstract and tangential relationship to software, or “IT”.

All of this was evident at a recent Open Knowledge Foundation meetup at the Wikimedia Foundation offices — but perhaps only evident to me, and I do not really intend to criticize anyone there. Their projects are all great. Nonetheless, I think very general discussion about open data tends to be very suboptimal, even among experts. Perhaps this just means general discussion is suboptimal, except as an excuse for socializing. But I am more comfortable enumerating peeves than I am socializing:

• “Open” and “data” should sometimes be considered separately. “Open” (as in anyone can use for any purpose, as opposed to facing possible legal threat from copyright, database, patent and other “owners”, even their own governments, and their enforcement apparatuses) is only an expensive policy choice if pursued at too low a level, where rational ignorance and a desire to maintain every form of control and conceivable revenue stream rule. Regardless of “open” policy, or lack thereof, any particular dataset might be worthwhile, or not. But this is the most minor of my annoyances. It is even counterproductive to consider, most of the time — due to the expense of overcoming rational ignorance about “open” policy, and of evaluating any particular dataset, it probably makes a lot of sense to bundle “open data” and agitate for as much data to be made available under as good of practices as possible, and manage expectations when necessary.
• To minimize the need to make expensive evaluations and compromises, open data needs to be cheap, preferably a side-effect of business as usual. Cheapness requires automation requires software requires open source software, otherwise “open data” institutions are themselves not transparent, are hostage to “enterprise software” companies, and are severely constrained in their ability to help each other, and to be helped by their publics. I don’t think an agitation approach is optimal (I recently attended an OpenOakland meeting, and one of the leaders said something like “we don’t hate proprietary software, but we do love open source”, which seems reasonable) but I am annoyed nevertheless by the lack of priority and acknowledgement given to software by “open data” (and even moreso, open access/content/education/etc) folk in general, strategic discussions (but, in action the Open Knowledge Foundation is better, having hatched valuable open source projects needed for open data). Computation rules all!
• A “data marketplace” should not be the first suggestion, or even metaphor, for how to realize value from open data — especially not in the offices of the Wikimedia Foundation. Instead, mass collaboration.
• Open data is neither necessary nor sufficient for better governance. Human institutions (inclusive of “public”, “private”, or any other categorization you like) have been well governed and atrociously governed throughout recorded history. Open data is just another mechanism that in some cases might make a bit of a difference. Another tool. But speaking of managing expectations, one should expect and demand good governance, or at least less atrocity, from our institutions, completely independent of open data!

“Nuance” is a vague joke in lieu of a useful title.