Annotation Systems
Annotation System-s typically allow any reader/user to attach a comment (structured or not) to any web object-node they find. https://en.wikipedia.org/wiki/Web_annotation
Systems that make Annotation visible at the original source
- modern: Hypothesis
- There have been browser plug-ins like Third Voice for these.
- Microsoft started to launch MsSmartTags, but an outcry caused them to pull the plug. The biggest issue was that they would be the primary supplier of annotations.
- There have been proxy servers like Crit Org http://crit.org/
W3C standards (2017) https://en.wikipedia.org/wiki/Web_annotation#Web_Annotation_standard
In the short-term it seems like a bad idea to have annotations spread through multiple non-integrated systems. On the other hand, maybe it makes sense to have each Virtual Community running its own annotation system. But how many communities are you part of?
- This seems similar to the issues of having Chandler/OSAF hold your personal notes, while also allow sharing/mixing of content in something like a shared Team Wiki.
dead: Annotea Semantic Web activity
I wonder whether BackLinks creates a similar effect...
Designing for "Control" of Public Annotations
Context: we would like shared annotations to be a tool for increasing collective intelligence.
Forces
- Some pages will draw 0 annotations, some will draw a smallish number, some will draw more than anyone normal can read.
- Some creators don't want any annotations or other interactive elements (e.g. blog comment) on/around their "content".
- Some creators only want non-critical (in-tribe) comments.
- Some creators want to "host" a "constructive" discussion, even if it includes critical comments. But they don't want trolls. Yes this is subjective.
- Some commenters use any channel available to "harrass" (via quality or quantity of comments) creators whose ideas they disagree with.
- Some readers want to focus on the "most popular/active" threads of discussion.
Solution:
- There will be many Annotation servers, services, formats/protocols. Most but not all servers/services will support the open standards.
- A commenter may have different identities on different servers/services (as with Mastodon).
- A creator can define a blacklist and a whitelist for hir page.
- A reader can define a blacklist and a whitelist for hir reading. This could be at the user-ID level, or the service/server level (as you can ignore certain Mastodon servers, regardless of user).
- Anyone can "subscribe" to other person's whitelists or blacklists, building a union-blacklist and union-whitelist.
- When viewing a page, a reader's browser will display: the gross number of annotations/comments on the page; the number of annotations/comments in hir whitelist and blacklist; and the number in the creator's whitelist and blacklist.
- The reader can switch between hir lists and the creator's lists. She can set a default for which choice to start with when viewing a page/server for the first time, and her browser will remember her latest choice for each page/server.
- This effectively defeats the goal of the creator wanting a "controlled" experience.
- A commenter can pick which service/ID to use when adding an annotation/comment.
- Some services will "rate" commenters for a reader based on WebOfTrust from commenters on that reader's whitelist/blacklist.
- A comment/annotation thread between a whitelisted and blacklisted commenter might initially "hide" the blacklisted comment, while making it easy for the reader to expose it.
- Is the reader/commenter's experience defined by the browser? By a browser extension? If the main browsers don't build it in, does it matter?
Edited: | Tweet this! | Search Twitter for discussion