Obelus Description

Obelus is mainly about web annotation.

• You might want to know what I am talking about. For you I have writen the basics
• You might know about it, and then the question is what is special about Obelus
• You might want to know why these names (Obelus,RdC)

The Basics

Roughly speaking Web Annotation is adding comments and information to web pages, without modifying them.
You can already leave comments on some sites (blogs, forums, newspapers), you can even comment on other people's comments. But not all sites have this built-in, and you could always ask for more features.
Therefore it would be interesting to have an independent server where annotations can be stored and retrieved, annotations with metadata (info on annotations' type and content) to allow features further beyond text comments.
In fact there has already been former research and developments about this stuff.
It is the recurrent purpouse of web annotation systems to be open (anyone annotating on any resource), and allow pointing the annotation to an specific spot (rather than the whole page) when applicable

Why is Obelus different

There are a few things I find unsatisfactory about previous intents, which I think they could be done better:

• The RDF-like annotations are cumbersome: I'd rather favour a basic fixed attribute set approach
• XPointer fragment locators seem clumsy and confined to XML resources. Regular expressions and tailor-made locators might do a better job
• Different versions of the same work might be available (e.g. translations). Annotations made on one of them may apply to the other versions (if one argument premise is false, it is regardless it is a English spoken video or a Spanish translated transcription).
• Previous implementations have earned the opposition of resource owners. Publishers should be granted some rights, so annotation is more acceptable, and they can even colaborate.
• A verbatim copy/excerpt of text resource might be made
  • because of localized annotation markup. In some content layouts (e.g. cumbersome dynamic pages) annotations might be difficult to display in the source resource. An excerpt of the text might be formated just as is suitable for annotations.
  • because of orphan annotations. source contents might be modified without notice, or become unavailable. It might be the publisher right to "erradicate" that content. But most times the publisher won't care enough about it.
• A framework for defining custom annotation types might be provided
• Annotation service available as a Web User Interface (otherwise we might depend on the client installing specific software)

About the names

Obelus(÷) is a text revision sign (see wikipedia:Obelus) mainly used to indicate that something is not OK about a material (e.g. if an ancient text attributed to some author or epoch contains anachronous words or expressions). Setting a warning signal on dubious information and giving adequate refutation would be quite an important feature on the internet(but also quite a disputable and abused one).

RdC (and rdcw) stand for "Red de Conceptos" (Network of Concepts). Initially the effort was in Spanish language, and one of the targets was the representation of knowledge as interwoven concepts (imagine each individual idea in a wikipedia page, and structure them as lists of facts about an issue, reasoning structures, causality, relations between questions, preferred/alternative answers/choices and why are they preferred/dismissed). I felt it was some kind of ethereal and unwieldy.