Thu 28 Nov 2019 12:38AM

Defining the "data commons"

P pospi Public Seen by 53

What do we mean when we say this? How can we best coordinate to build this collaboratively and emergently?


jean m russell Tue 10 Dec 2019 12:04AM

I mean having data that we share and using an explicitly "commoning" approach to that sharing. Personally, that means Ostrom-inspired principles but I am open to something broader. Specifically I want to be pointing to the benefit gained from creating context and meaning with others.


pospi Sat 14 Dec 2019 5:39AM

I really like championing Ostrom's work as a primary framing for how we design commons. I feel good about the idea of "commoning" as the broader cultural context. It is non-threatening, mildly partisan, but partisan in a pro-environmental way that is important right now. May be important to highlight (reflecting on "naming" thread) that there are multiple themes and narratives that we can weave together in this work; "commoning" and "sovereignty" are deeply inter-related.

What interests me are the technical and social specifics of how we create that cultural context- what does "commoning" look like in terms of the ways we design protocols? The ways in which data flows within a shared network space? In terms of the sovereignty of and agency over the actual bits on disk?


jean m russell Tue 10 Dec 2019 12:05AM

Also note, originally we were talking about supply chain commons, and I found that both too broad (how much of a supply chain is enough to call it a supply chain commons) and too narrow (can these tools be used if the data is not within a supply chain).


pospi Sat 14 Dec 2019 5:34AM

@Samuel Rose @Jean Russell this discussion in VF seems relevant to our most recent (yet somewhat long ago) touchpoint on data commons work - https://github.com/valueflows/valueflows/issues/603


Noah Thorp Sat 14 Dec 2019 6:24PM

Yes and I think we need to get more specific. Are we including software code as data? What makes data useful - ie usable formats? Who are our first commoners who have data? So far we mainly have code and not the other kind of “data commoners” - aka “users” ;-)