Contents in Current Page:
General Links are the other fundamental construction unit of hypermedia besides nodes. Usually an explanation label is attached to each link. This label is usually a free-text field and provides an explanation about that specific link. In general, a hypermedia design should explain to the user why the destination for a link was an interesting place to move to by relating it to the departure point. links in most hypermedia systems usually interconnect nodes in just one direction. Bidirectional links although relative easy to be implemented, are rarely used. In most cases links are predefined by the author of a hypermedia application (Alschuler, 1990). In general we can describe links in average current hypermedia systems as binary, usually single-directional, labelled but not typed, explicitly defined, hardwired interconnections between nodes.
Hierarchical Links Most links are explicit, in the sense that they have been defined by hypermedia authors to connect one specific departure node with one specific destination node. Except this well-understood type of explicit links, some hypermedia systems also provide hierarchical "semi-explicit" links. These type of links are implicitly defined in hierarchical structures between nodes. Links provided by hierarchical structures although are predefined, are different from reference links since users don't activate them using the regular link activation mechanism applied to reference links.
Virtual or computed links are links that are determined dynamically instead of being static predefined in advance by the hypermedia author (Kibby and Mayes 1989). For example, in a hypermedia system supporting computed links one can define the destination of a link as "the node containing the best picture for supporting user request X". Implementing virtual links has been proved a very difficult task for hypermedia designers.
One possible enhancement to the simple notion of link to basic hypermedia model is to allow for links having more than one destination. On activation of such a link the system could provide the user with the possible destinations and ask to select one. An alternative way to handle multi-destination links is to open all the destination nodes. Another alternative is to open the node which seems to match better the current user needs.
A link can be activated just by clicking in the associated anchor. Printing an anchor with boldface type or to highlight it, is a very common means to indicate a link to the hypermedia user. Another alternative is for the system to provide the user with pull down menus including all the links for the current node. Then, the user can select one link to activate.
Links - Span to Span Links cannot exist only between different nodes. They can also exist between two different parts of the same node. Furthermore a destination of link cannot only be a node as whole entity, it can be also a specific part of the destination node. This kind of links have implemented in few hypermedia systems and are named span-to-span links (Halasz, 1994).
Links - attached procedures Links in the basic hypermedia model can be extended to support attached procedures. An attached procedure is activated when the user activates the link (Rada, 1991A). The graph which is extenuated to include procedures is called an augmented transition network and it has proved by mathematicians that it can represent any computable algorithm. We can understand that a hypermedia network enhanced with attached procedures could be a very expressive and powerful information environment.