On 9 January 2013 13:13, Michael Rogers
<address@hidden> wrote:
> The first use case should be, can I add a friend from system A to system B. We need to ask projects how they identify their users in digital string form, and if it's possible to translate that string of characters to something another system can understand unambiguously. If yes, we're good. If not, move to the next.
I'm not sure that being able to add friends across systems is the same thing as being able to identify users in string form across systems.
For example, Briar has no global user identifiers. When two people add each other as contacts (which takes place face-to-face or via a mutually trusted third party, so there's no need for a global identifier), each person chooses a nickname for the other. Those nicknames only have local scope.
In principle, two Briar-like systems could interoperate and allow users to add friends across systems, without any way to identify users using globally unique strings.
Sure that sounds like a very reasonable way to go. Many systems do not have a string that identifies a user at all.
My approach was more aimed at interop with the wider web. Just as you can link from document to document if you are on the web. You can friend from system to system if that system has an identifer, and you know what it was. Millions are doing this already today, and networks get more fun as they grow.
Cheers,
Michael