Table of Contents
Index
Glossary
Table of Contents
Index
Glossary

  
The readers' aids

A modular article in an electronic environment may have many complicated features. The reader therefore has to be able to customise the presentation, which implies that the most important goal of the readers' aids is flexibility.

Locating, retrieving and printing 

Readers require a search engine that allows for complex searches of entire articles,  complex modules and  elementary modules. It must be possible to search by:

• module labels;
• phrases from the full text in any particular module;
• link labels;

or combinations thereof. For instance, the reader must be able to locate a microscopic Results module, dealing with the differential cross section of collisions between sodium and iodine atoms, that is connected to a module Experimental methods, about a molecular beam set-up, by means of a problem-solving dependency link. Another example is the  search for a mesoscopic Experimental methods module on surface ionisation, that is the target of at least ten links carrying proximity-based labels indicating that the source is not part of the same article and resolutions-labels indicating that the target provides details.

Once the reader has located the modules of interest, he should be able to print individual modules (automatically including the images and other non-textual but printable parts) as well as any selected collection of modules.

Modules

Just as for the traditional scientific article, the modular article, as well as each individual module, is a self-contained representation of scientific information. Therefore, we must keep in mind that many of the same presentation issues that have to be taken into account when writing a traditional article remain valid in the modular case. The internal structure of modules has to be made visible by means of typography, paragraphs and sections. If the reader prefers to print the module on paper, the presentation of the printed version may be adapted to the paper medium.

In addition to the traditional presentation requirements, the modular structure leads to requirements concerning the composition of modules and links.  Complex modules, which consist of linked constituent modules and a  `module summary' summarising them, should be implemented in a `recognisable' way. For example, the module summary may be presented in a different font or colour than the  elementary modules. The exact presentation should not be hardwired into the module, but rather stored in SGML type tags. The presentation then can be managed through the style sheet.

The reader's main requirement concerning the implementation is that the presentation should be flexible. In particular, the reader must be able to unfold or hide:

• mathematical digressions and other details;
• parts of modules that overlap with other modules that the reader may already have consulted;
• the characterisation of the module and the navigation menu;
• full figures (which can be replaced by thumbnail figures);
• unprintable representations of information (e.g. java applets).

Links 

The distributed presentation obscures the coherence of the information: if the reader does not understand the target and the nature of the links connecting the modules, he cannot make a well-considered choice as to whether to follow the links or not. Therefore, the type of the link must be made explicit as well as its target. Rather than identifying the target with an uninformative identification code, the author name and publication date of the cited module may be made explicit. In addition to the type and the target of the link, a short phrase can provide further clarification. This characterisation should be hidden from view, and only be made visible on demand. For example, the link may be represented in the text by a small icon (different icons may be used to indicate the main function of the link) and the characterisation may be shown in a `pop-up' box when the reader moves his mouse onto the icon.

If a particular point in a module serves as a starting point for more than one link, the various links may be routed via a menu. For example, at a particular feature in the graphical of the results a link may be provided for zooming in on that feature, another for comparison of that feature to a similar feature in another Results module, and a third to its interpretation.

Many explicit links could also make a modular article unreadable. The reader must therefore be able to choose how the links are presented: as elaborate informative icons, as unobtrusive icons, as words, or completely hidden from view. The different types of links should in principle be distinguishable at first glance. Links that have been created to express an organisational relation may be presented using a different colour, font or icon than links created for the different kinds of scientific discourse relations.

Furthermore, the reader requires tools that enable him to:

• customise the organisation of his screen. For example, if he activates a link, he can view the target:
  • in the same window that the source of the link occupied previously;
  • in a separate window overlapping the window presenting the source;
  • in the other half of a `split-window' that allows for simultaneous viewing.
  The organisation of the windows may depend on the type of the link. An external link may, for instance, be considered a detour and lead to a separate window, whereas a link expressing a similarity relation calls for direct comparison and thereby for two neighbouring windows; a sequential link would stay within the same window, and the Map of contents may be permanently kept in a corner of the screen.

• define a personal sequential path. In the modular model, we have defined sequential relations that allow the reader to consult the complete article (via the complete sequential path) and to consult the modular article as if it were a traditional, linear article (via the essay-type sequential path). The reader must also be able to define a personal route through the article, which he then can follow by way of `next' buttons. Navigation links for those paths should be provided at the bottom of a module, allowing the reader to simply `turn the page' after reading the module, but also at the top to avoid forcing the reader to scroll through the entire module.

Overview  

At all times, the reader of a modular article should have at his disposal the navigation menu of the module at hand, as well as the Map of contents and the Abstract of the article.

In an electronic environment, the map of contents must be interactive: the representations of the modules in the map must be linked to the modules themselves by hyperlinks. The map should also indicate (on demand) which `content module' the reader is currently consulting and which modules he has accessed before (if any). The map of contents must be flexible, because it is too complex to be presented in full detail to the reader. The reader must be able to:

• choose which types of relations are visible and to what degree of detail he wants to view the map;

• zoom in on the details of modules and links. The default presentation includes each module in the article without links, except the complete sequential path. The reader may also hide all constituent modules and concentrate on the main modules, or add more details such as (all or some specified types of) links between (all or some specified) modules, the mesoscopic modules linked to the modules of the article, or the full physics characterisation of modules.

• generate a Map of contents representing the structure of a larger network of modular publications based on the information stored in the  `journal database' about the modules and the links connecting the various modules.