Re: Transfer of followers - a very interesting question and a major use case for DID standard.

Markus, I beg to differ that this use case is out of scope for DID standard.

Before I go to explain why, I have a proposal to classify DID use cases
like this:

   1. Any use case where DID can be used is a valid use case for DID spec =
   **Minimal DID use case**
   2. Use case that has broad applicability, considerable impact on the
   Internet + Point 1 = **Good DID use case**
   3. Use cases where DIDs usage provides more security, privacy over CIDs
   (Centralized Identifiers)  usage + Point 2 = **Great DID use case**
   4. Use case  where DIDs usage enables new, better functionality over
   CIDs + Point 2 + Point 3 = **Awesome DID use case**
   5. Use case that is applicable in the current real world + Point 2 +
   Point 3 + Point 4  = **Majestic DID use case**

"Transfer of Followers" use case definitely meets point 1 and 2. I also
think that this use case indirectly meets the 3rd criteria.
As I put in the paragraph about "Real World Dimension" in the first email,
most people use "Base Identity" to log in into different services. Hence,
this Base Identity is very important, hence it should be Self-Sovereign,
hence it should be based on DIDs.
One can say, that this use case is a derivative of Base Identity use case,
but because of its importance, I think it should be described independently.


I think your ideas that the list of followers should be independent of
a specific
> service, and that your followers (or social graph) should be fully
> portable, align very well with SSI.


One, clarification to this - I think that followers should be portable, but
I do not think that followers should be independent from a specific
service. I am not sure if we need to use lists of followers in the real
world. Followers are context specific. My followers on SoundCloud are not
the same as my followers on Twitter. As a user of SounCloud, I want to have
a simple way to say to my followers on there - "Hey, follow me on Spotify
now. Here is a button to do this in 1-click.". I do not think it is a good
idea to handle this use case by creating Lists of Followers and somehow
contacting them independently of services, where these followers follow me.
This would require new UX, local applications >> too big mental load >> not
practical in the real world.

My proposal to handle "Transfer of Followers" use case is to enable
functionality to invite followers from one service to another,  in the real
world within the current Internet services as easy as possible.

In the real world, on the example of Spotify and Soundcloud, it will mean -
Spotify should have legally enforced ability to have access to SoundCloud
API, in order to be able to provide me functionality to invite all my
SoundCloud followers to Spotify. When I make this invitation on Spotify
(obviously Spotify will be very interested to provide this functionality to
me) all my followers on SoundCloud should receive clear message "Hey I am
on Spotify now. Follow me there. Here is 1-click button to do this.". To
make such invitations as easy as possible, we will probably need to do it
by employing Base Identities which people use to log in everywhere, anyway.

To some extent this use case can even meet criteria 4 and 5 ("to some
extent" because this use case can be achieved with CIDs and to make it
feasible we need new competition law for this), making it almost a
**Majestic DID use case**.


Bohdan

P.S. Not sure, if my selection of adjectives to classify DID use cases is
W3C compatible))) Feel free to propose better names.



On Tue, Jul 3, 2018 at 5:42 PM, Markus Sabadello <markus@danubetech.com>
wrote:

> Hey Bohdan,
>
> This is out of scope for the DID spec itself, but definitely an
> important use case that can be built on top of DIDs.
>
> I think your ideas that the list of followers should be independent of a
> specific service, and that your followers (or social graph) should be
> fully portable, align very well with SSI.
>
> W3C ActivityPub has a way to express in RDF/JSON-LD who your "followers"
> are, and who you are "following", see here:
> https://www.w3.org/TR/activitypub/#followers
>
> Markus
>
> On 06/30/2018 02:27 PM, Bohdan Andriyiv wrote:
> > On the last CCG call, Christopher Allen raised a very interesting
> > question – transfer of followers from one service to another. This is
> > a very interesting and important question, as well as a major use case
> > for DID standard. If such transfer of followers can be done, it would
> > have a profound impact on the Internet. Every Internet industry that
> > have social networking element, will become much more competitive as
> > incumbents will lose their lock-in on users, caused by the network
> > effect.
> >
> > Now, when we see, that making transfer of followers possible would
> > have a huge positive impact on the Internet, the question is - "Is
> > such transfer of followers a theoretical possibility  a
> > "would-be-really-nice-to-have-but-never-gonna-happen-thing" or is it
> > something that can be done in practice and if yes, how?".
> >
> > To answer this question we need to look in 3 dimensions.
> >  * Technical dimension
> > We need to have Internet-wide accepted standard of identity. Such
> > standard should allow me to say - "I am identity "realBob123" on
> > service X." and every service should understand it. If this can be
> > done than I will be able to say something like this  - "Hey service A,
> > I do not like you anymore, I am using service B now and I am identity
> > "realBob123" there. Show on my profile page a button with text "Follow
> > me on service B", and also send message to all my followers with this
> > button. And when my followers click on the button make sure they will
> > register and automatically follow me on service B."
> > BTW, maybe I am rediscovering America here. Do we have such
> > inter-operable standard of identity already? Can DID become such
> > identity standard or it is already covered by other standard?
> >  * Legal dimension
> > Transfer of followers functionality should be enforced by competition
> > law. Just as Google Chrome browser, allows to change its search
> > engine, all Internet services that have social component have to allow
> > simple standard way to tell your followers - "Hey I am on another
> > service, here's a button, click on it and you will start following me
> > on that service.".
> > Now, thinking about it, I am really surprised this has not been
> > proposed already, after all of the negative press with Facebook.
> >  * Real world dimension
> > Besides network effects, real world behaviors are shaped by user's
> > need for minimize mental load and habits.
> > To achieve this mental load minimization, we would need to have a
> > standard way on all services to initiate followers transfer. There
> > would need to be a /standard/ form on all services with fields: 1) New
> > service where you want to be followed; 2) Your identity on a new
> > service. Such form would need to be enforced by competition law.
> > But, unfortunately, this form might still be too complex – too high
> > mental load barrier for many people in our hectic world.
> > I think, to make followers transfer really seamless and commonplace in
> > the real world we would need to use **Base Identity**.
> > In the current Internet, most people use 1-click sign up/in usually
> > with Facebook or Google identity to most of the services. This,
> > Facebook or Google identity is what I call a "Base Identity". When
> > this Base Identity exist, and in real world settings it does exist for
> > most people, the transfer of followers can be intermediated through
> > it. A user will not have to fill in a form to tell Service A his
> > identity name on Service B, he will just have to say - "Hey followers
> > on Service A, follow me on Service B". What is important, the
> > initiation of followers transfer, now will be possible from service B.
> > Service B will be able to provide to a user a functionality to offer
> > his followers on service A to follow him on Service B. As Service B is
> > very much incentivized to provide such functionality, it will figure
> > out the best seamless and easy design to do it - hence no need for
> > hard to understand forms on service A (which Service A is going to
> > hide anyway as it is against its incentives).
> > Such Base-Identity-intermediated transfer of followers, relies on the
> > existence of Base Identity, and subsequently Base Identity provider.
> > This immediately, opens a number of questions: "What exactly is Base
> > Identity", "Does Base Identity exist in the current world?", "Do we
> > really need Base Identity?", "Is it not too dangerous to have Base
> > Identity?", "Who is or should be Base Identity Provider?", "Would Base
> > Identity Provider have too much power?", "Can Base Identity be
> > self-sovereign?". I would very much like to explore and discuss Base
> > Identity questions, probably in another topic.
> >
> > Conclusion.
> > Transfer of followers is very important. It will make the Internet
> > better, by making it more competitive.
> > Transfer of followers is possible in practice. It will require
> > technical standard and enforcement by law.
> > In real world, transfer of followers is going to be intermediated via
> > Base Identity.
> >
> > Bonus: Base Identity concept. Base Identity is a very important real
> > world existing concept. Base Identity and Base Identity Provider
> > concepts and their aspects (existence, importance, relevance to DID
> > standard) needs to be explored more.
> >
>
>
>
>
>