In five posts, I want to speculate about the next twenty years of open data based on the past twenty years of open source. The idea is to transfer what we learned from open source in one way or another to open data.
This is part 2 on using open data.
Reciprocal licenses and FUD
Enemies of open source were able to stall its progress by creating FUD (fear, uncertainty, and doubt) around reciprocal (“viral”) licenses. The same might happen to open data, where share-alike (reciprocal) licenses also exist. Companies might worry about using open data, because they fear it might poison their data, requiring them to share their own data as open data against their will.
On a more practical level, however, it is usually the absence of an open data license for some publicly available data that stalls the growth of open data.
Governance and license compliance
Any company that knows what it is doing has some sort of open source program office (OSPO) that governs the use of open source. I expect to see analogous ODPOs whose task it its to educate the company about open data, clear its use (licenses), and support developers in delivering applications that fulfill the license obligations.
Rest assured that for the next twenty years any open data conference will have a track on open data licenses and how to comply with them.
Liability for the future
Old products often contain undeclared open source code, which creates problems during M&A due diligence or when an open source monetizer gets their binary analysis tool going. The lack of open data governance today may well create a liability for the future, consisting of contaminated open data projects and proprietary databases alike.
Separately from this, using open data will have lots of original challenges, most notably around quality, provenance, etc. that may sound similar to open source but probably will require solutions unique to open data.
Next up: Contributing to open data projects.