Suche Home Einstellungen Anmelden Hilfe  

prev next Up Title Contents Index

1.2.2 Link Management

Hyperwave Information Server does not store link information directly within the source document, as is the case with most classic hypertext systems. In these cases, because the link information is only available in the source document, it is difficult to maintain the link consistency of (large) Web sites. When you want to delete a document and need to know what other documents refer to this document, you have to search through all other documents for hyperlinks to this document. Obviously, this is prohibitively expensive for an intranet site with tens of thousands of documents.

LINKS SEPARATED FROM DOCUMENTS

To avoid these problems with link consistency, Hyperwave Information Server has a very different approach to link maintenance: hyperlinks are separated from documents and stored as objects in their own right, in the same object repository as documents, collections, etc. In the case of HTML pages, for example, this means that the hyperlink information is extracted when the HTML page is inserted, stored in the repository, and dynamically merged back into the document when it is served to the user.

Links are realized as one or two objects: a source anchor defines the region the link is emanating from (e.g., a sequence of characters in a text or a region of an image). The source anchor points to either

Because links in the Hyperwave Information Server are not statically stored, as in classic hypertext data models ( Figure 3), many traditional problems are overcome, and many extra advantages gained. These advantages are listed below:

Figure 3 :The Classic Hypertext Data Model


prev next Up Title Contents Index

Benutzer: gast • Besitzer: hwsystem • Zuletzt geändert am: