From Federated Social Web Incubator Group
The Social Web Acid Test - Level 0 is an integration use case for the federated social web.
The use case
- With his phone, Dave takes a photo of Tantek and uploads it using a service
- Dave tags the photo with Tantek
- Tantek gets a notification on another service that he's been tagged in a photo
- Evan, who is subscribed to Dave, sees the photo on yet another service
- Evan comments on the photo
- David and Tantek receive notifications that Evan has commented on the photo
Original detail (that may have been lost in moving to this wiki):
Note that the first two steps above are actually part of the same action:
- With his phone, Dave takes a photo of Tantek, tags it explicitly with Tantek, and uploads it using a service
- Dave, Tantek, and Evan have accounts on different social networks
- Corollary: either they don't have accounts on the same services, or if they do, there is *no* dependence on them having their accounts on those different services "connected" (analogy: just like you can send email to other services without having accounts on those other services, you should be able to receive notification of being tagged, or of an update, or of a comment, without having to have an account on the same system as the tagger, the updater, the commenter).
- These social networks have at least two different codebases
- There is no (spoon?)
- All interactions between social networks are according to published open standards.
- Zero dependency on proprietary (AKA snowflake) APIs/protocols/formats
- Zero dependency on any biz-dev or any other a priori sales/marketing relationship/negotiation between the sites.
- user A takes a photo of B from their phone and posts it
- user A explicitly tags the photo with B
- B gets notified that they are in a photo
- C who follows A gets the photo
- C makes a comment on the photo
- A and B get notified of the comment
Where users are on at least 2 (ideally 3) different services each of which is built with a different code base, and the users only need to have *one* account, on the specific service of their choice (requiring the users to have an account on each service that is interacting misses the point of Federation)
And (ideally? or we require that) participants A,B,C, use their own sites (personal URLs) running their code (see throwdown).
Originally from http://tantek.com/2010/199/t3/fsws-social-web-acid-test-swat-v0
One particular detail (that may have been lost in moving to this wiki:
Note that the first two steps above are actually part of the same first action:
- user A takes a photo of B from their phone, explicitly tags the photo with B, and posts it
- B gets notified that they are in a photo
extracted from: "user A takes a photo of B from their phone and posts it, B gets notified that they are in a photo, ..."
The two step variant (post, *then* tag) is interesting, but arguably not as typical a real-world case, as was the intent behind David Recordon's explicit suggestion of using a *phone* to take and post the photo with tags.
SWAT0 Open Standards Protocols and Formats
Feel free to offer strawman / hand-wavy thoughts on what protocols and/or formats Federated Social Web implementations may use to interoperate the steps in SWAT0.
But if you add a buzzword, describe exactly what function it would perform in which concrete/specific step or steps in the scenarios described in the SWAT0.
- ... technology buzzword 1 ...
- ... enables implementing step n in SWAT 0 by providing a protocol for user A's site to (do something) to user B's site. (etc.) ...
In order to keep the page cleaner (and allow other to post other strawman), here is a first shot at it: SWAT0 - strawman protocol flow
Here's a strawman of a variant of SWAT0 using only textual messages (i.e., no photo tagging) that should have similar functional and technical requirements.
Frequently asked questions about the SWAT 0.
- Q. My software doesn't support sharing photos/tagging photos.
- A. It's going to be hard for you to participate then. It's totally reasonable for you to formulate a similar test and starting working towards it.
- Q. SWAT1 should require shooting a video while riding a horse and sharing that video with a government official on the moon! Who is asleep!
- A. We're trying to focus on getting SWAT0 working by September 30, 2010. Based on our experience with this implementation, we'll try to formulate what SWAT1 will look like (if it happens). Trying to define it now is a good way never to make SWAT0 go.
- Q. Do the solutions have to use OStatus?
- A. No. Solutions in XMPP, JSON, XDI, RSSCloud, or whatever are totally acceptable.