These remarks cover only items creted before July 2000, numbered 1-8 in the April 2001 menu.


Be sure to adjust your screen display and font sizes to maintain correct legibility. Internet Explorer 4.0 and higher are required for full functionality; Netscape browsers will not manage all the features.


A condensed version of the Main Menu will open automatically for persistent remote control.

The advantages are 1) rapid switching becomes possible among the various demos (without returning to the Main Menu); 2) returning to the Main Menu will also launch it again; 3) the window can be repositioned or resized.

The disadvantages are 1) the popup will conceal what lies beneath it, not always predictably; 2) when clicking away from the remote control window, it will lose focus and become hidden, requiring relaunch from the tab bar; 3) when minimized it cannot be relaunched from the Main Menu; 4) local font sizes in effect may make the popup window hard to read.


Plain text following a printed edition by using BREAK tags after each line.

Page BGCOLOR=white TABLE align=center border=20 bordercolor=green cellspacing=10 cellpadding=20 TH BGCOLOR=ivory

TD valign=top BGCOLOr=ivory


Same as [1] with line numbers in another TABLE cell, synchronized line-by-line through the use of the same FONT FACE and SIZE as the text cell but highlighted in another color. Instead of a green TABLE BORDER of 20 pixels the entire page BGCOLOR is green.


This is an earlier version of [2] with more complex FRAMES coding to achive a similar result. The separated scrollbar, an accidental result, is preserved here for its interesting design.


Context information and pull-down menus. While it is possible to do small-scale coding of a demonstration paragraph or two by adding extensive commentary links to those text, a different procedure would be required to do large-scale coding of all of ULYSSES. One approach would be the on-the-fly method to create a display of a specified text with status information and annotation menus -- as this dummy page should suggest. Let's assume that all of the project data is in a huge relational database in which each of dozens of potential forms of information in its own table, all linked by positional information, probably a unique index number referring to the episode/section/paragraph/feature. (There is no confusion as to the eighteen episodes or chapters of ULYSSES. A section would be a group of paragraphs or lines to which shared commentary applies. Paragraph numbers could be assigned which would be reasonably stable from edition to edition. The features to be annotated would often be located within a paragraph but might also be relevant to a larger unit, such as one or more sections or episodes, or it might be a word or phrase which appears once. Because the database would be "aware" of recurring text, topics, or annotations, it might provide away of cross-referencing or cross-navigating as well as a way of reducing encoding duplication. )

By displaying text "on demand" the database manager would "know" what is on the screen and could therefore display relevant status data (episode name, paragraph numbers, characters present, scene location, time of day) and appropriate menus for additional information (motifs, annotation, controversies, rhetoric, criticism, text variants, corrrespondences, multimedia, etc.). The representation given here is conceptual; in practice the reader or user would be able to select or customize a configuration which might correspond to approaches (Gilbert, Ellmann, Kenner, Derrida, etc.) or to expertise (student, teacher, researcher, scholar), or interests (see examples and imagine your own).

DEMOS 5, 6, 7.

Four annotation panels, three above the text and one below it, using FRAMES. These working demos consider various ways of linking a text to annotations on it (in this case, Gifford, Schutte, my footnotes, and a concordance). One demo [5] provides labels and fixes the size of the various panels, each of which can be scrolled. A second demo [6] disposes of the labels, enlarging the panels and permitting them to be resized. A third demo [7] locates the links at the line numbers so that multiple annotations can be retrieved at once through a JavaScript function.

In a larger text and annotation sample and with more programming time, secondary links would be useful to follow up on other instances of interesting words and motifs in the text. The assumption that screen space is fixed is probably a useful one for the beginning reader of a hypertext, but in more advanced uses multiple layers, popups, and other devices can be used once there is some assurance that the ready would not be lost or disoriented.

DEMO 8. Mouseovers.

(Internet Explorer 4.0 or higher recommended; limited functionality in NetScape Navigator).

Visual clues on the screen (the standard blue underlines) invite exploration by the mouse, which will produce graduated levels of information.

STATUS LINE: Helps on the bottom line of the screen, limited in size to one line, not especially prominent.

TOOLTIP: Short display (about five seconds) of a balloon popup similar to ALT messages for images. Can be displayed again by re-positioning the mouse. Can be long enough to preview what lies ahead for left-clicking.

CLICKING: Change display in right pane to entended commentary.

HIDE/DISPLAY: Can be a quite large and elaborately formatted text which will stay in place until mouse is moved away. Awkwardly, will be limited to space available between mouse position and bottom of screen, requiring a second try after adjusting the text upward when the popup display will not fit.

PERSISTENT POPUP WINDOW: Unlike other Mouseovers seen on this page, this popup window will persist until closed. It is useful for information that is more than transitory but can be inconvenient if the user fails to close it after use or inadvertently opens multiple instances (which must be cleaned up on closing the session.)