[License-review] Request for approval by license steward: Tidepool Open Access to Health Data Software License

[Josh Berkus]

On 10/07/2013 08:00 AM, Howard Look wrote:
> Thanks John, that's a great point.
> 
> The Tidepool license doesn't require any specific technical means
> for machine-accessibility, though we definitely would like to
> encourage people to provide (say) RESTful APIs over CSV downloads
> since that's the most likely mechanism for encouraging openness
> of data for a patient and accessibility to other programs. That said,
> we also want to keep it technology agnostic.
> 
> A couple of possible options:
> 1) We try to come up with some language that is technology agnostic
> but requires something like "machine-accessible data access in both bulk
> and individual data point form." Would that be considered technology agnostic?
> 
> 2) Leave the license language as is, but change the FAQ to encourage,
> but not require, RESTful APIs over CSV download. This wouldn't be terrible
> (since someone can always write a CSV to RESTapi 'converter') but it would
> be a bummer since it would tend to inhibit patient-initiated data exchange.

Speaking as a developer, it would take me all of 20 minutes to write a
CSV-to-REST converter.  Maybe an hour.  And then I'd be happy to
open-source it because really, how hard was it to write?

The format of the data received is the wrong thing to focus on.
Instead, you should focus on the issue of "continual" vs. "one-time"
access. CSVs are a *great* format with lots of library support.  The
critical issue is whether I can get my insulin data from this morning,
or whether I have to register on a 3rd-party website to get a CSV dump
once a year and sign up for a spammer list in the process.

There is also the issue of whether or not the data supplied is
*documented*.  You can give me a nice whiz-bang REST API, but if that
API is only documented in papers not available to the public, then some
data analyst is going to have to spend a couple of weeks figuring it out.

Some sample language:

"The User shall have the right and ability to access their complete,
up-to-date, documented, and machine-readable structured data, on demand,
as many times as the User requires.  No additional requirements, whether
fee, registration, or other burdens, shall be placed on the User for
access of the User's data except as absolutely necessary to verify their
identity."

Then in the FAQ you can add an explanation and examples of what
"machine-readable structured data" is.

--Josh Berkus