This page is for brainstorming about ideas, proposals, constraints, requirements for an Attention microformat.
This is part of an open standards community effort to update (even rethink and rewrite) the Attention.xml format on a strong foundation of research of existing examples and formats, as well as leveraging the Attention.XML implementation experience from Attensa, CommerceNet, Technorati, and any other interested Attention developers.
- Tim Brown
- Tantek Çelik
- Eric Hayes
- Rohit Khare
- Kevin Marks
Both Technorati and Attensa have implementation experience with Attention.XML and other attention formats.
Other Attention-like Things
- Hierarchy in the format gets in the way more than it helps
- Need to define a minimally useful subset for ease of implementation and interoperability
- Need to "chunk" any sets of additional data
Goals / Requirements
- Needs to be represent sparse (over time and source) data
- "Blendable" - get an attention "chunk" every day from a source and be able to aggregate the chunks into one attention file.
- Flat by default. Permit hierarchy but don't require it, and don't require parsing of a hierarchy (i.e. allow treating the attention as a simple flat set of attention items).
These are things we want to avoid in an Attention format, mostly due to privacy reasons.
- login information
- personally identifiable information in URLs (e.g. PHP session info)
Strawman Schema Thoughts
Each Attention Item contains
- item url (rss article url, etc)
- parent context (feed or source site, )
- attention date (article last read time)
- duration (attention duration, how long did i look at this page or article)
- followed links (click through. could result in a next attention item)
- read state
- date removed