Hi everyone,

Splitting the thread off to avoid hijacking it

On Mon, Jan 18, 2016 at 9:50 AM, Mitar mmitar@gmail.com wrote:

I think this conversation is diverging from the question of the *service* we should offer to others to licensing of the content. Licensing does not say anything about the service one should offer for the content. Any service, any API, is more or less something one does extra on top of the licensing requirements. We could just offer dumps of data and this is it. But if we offer more, some specialized services, uptime and availability and so on, that does not have much with the licensing of the content. That discussion should thus be on some other layer. Investigating licensing will not give us much insight into the question if we should go into the business of offering data services or not.

I think this is a useful way of thinking about the problem. One thing we discussed quite a bit at the Wikimedia Developer Summit earlier this month is the distinction between the content format (see "content format" < https://phabricator.wikimedia.org/T119022%3E) and the APIs that we use to access the content (see "content access": < https://phabricator.wikimedia.org/T119029%3E).

The two are incredibly easy to conflate, in part because one could argue that the content format is merely a translatable expression of the underlying data model. That said, it seems to me that we have to stop abstracting things *somewhere*, to avoid getting deeply lost in too many layers of abstraction. If nothing else, we need a "free format" per the Free Content definition (http://freedomdefined.org/).

Mitar, is your layer distinction between "service" and "content" the same one that I'm trying to draw between "content format" and "content access"? I have further thoughts on this, but I just want to make sure we're talking about the same distinction.

Rob