The reason why interacting with content could be interesting is explained by the example of the current article. The concept maps describing this article will have sections of the article as content. The text in these sections will contain hypertext links to other parts of the article. But when using these hyperlinks, the context described in Conzilla will not automatically be updated to show the context of the link end. Thus, it would be desirable to be able to either automatically update the context when following a link, or to have a button such as "Show context" in the article display that shows the context of the current text. This would demand that Conzilla be controllable using, for example, JavaScript.
It should be clear that allowing this type of manipulation of context from inside of content could be immensely useful when navigating complexly interlinked content collections, which is probably not at all going to be an unusual case.
The requirements on Conzilla for it to be able to interact with external tools include:
It has to have a standardized interface. Note that the specification of this interface may very well have be made completely concept browser independent, as content is not supposed to know which browser the user is using. This places strong demands on the interface, which must be carefully designed to meet the following criteria:
There must be a means to locate Conzilla and the interface. This includes specifically from inside of content, where at all possible.