Open Hypermedia Systems Working Group

Fred the programmer - Analysis

Preliminaries

Author:
Kasper Østerbye

U Aalborg, Denmark


Updates
Most recent: 13 March 1997


References


Other Links


• Scenario
  
• Analysis
  
• OHS mailing list discussion (when available)

Analysis of the scenario:

There is a need to be able to address nodes across hypertext systems. As an example if one wants to link from documentation maintained by the CCLE to a function which is maintained in a file managed by the programming environment. My thoughts in this direction is that from the point of view of the CCLE, the programming environment is a huge composite, and it is possible to address nodes and links in that composite, perhaps in a composite specific way. I thus believe that composites can play a vital role in cross-hypertext integration. This will likely produce new and interesting variations on dangling links, and perhaps also "dangling composites". Notice that the CCLE also need to export itself as a composite, to allow the reverse integration, linking from the programming environment to the CCLE.

As I should have expected the versioning and configuration issues always play an important role. Though several persons myself included have come up with flexible and powerful data models that can handle the issues presented in the scenario, I now believe that the real challenge in versioning and configuration is to develop a user model that is simple enough to be useful in practical situations. Cross hypertext linking does complicate the matters further and I have no solutions.

Also notice that the hypertext system is "at the core". There is hypertext functionality everywhere at the author level as well as the reader level. Eg., it is easy for Wilma to create a few free hanging notes and links and send them by e-mail to Fred. How this is done I have no clue. But the separation between loc. specs and ref. specs as advocated by Kaj and Randy might be all that is needed to make emailing hypertext practical.

Goals of the Scenario

Fred has a bad day


The overall goal is to investigate through a scenario where I assume the existence of the kind of open hypermedia systems we are looking at today, the kind of problems that will arise for various practical reasons. The goal is to see if OHS will solve the real problems.

User models


Some applications will have their own internal hypertext systems, help systems, or as in the scenario, where the programming environment offers computed links to navigate the source code. How do we make a consistent user-interface across a number of hypertext systems?

Integration of different hypertext systems


Normally integration has been taken to mean usage of 3rd party editors for node contents. When we are successful in getting OHS systems on the market we must also address how these can themselves be integrated. At least we must consider how we can combine internet and intranet hypertext, where the important issue is who owns what.

Characters

Data

Requirements for Participating Applications:

Requirements for Link Server:

Requirements for Stores:

Implementations:

This section of the scenario is not to be filled in until the scenario has been circulated among the OHS mailing list (ohs@csdl.tamu.edu).

Protocol Requirements: