This paper offers advice and hints on structuring a Web document and customising the brower to remove the worse aspects of many slide presentations. These ideas are based on using the Web for presentation of lectures at the University of Canberra, and for presentation of tutorials at various conferences.
There is nothing particularly original about the techniques presented. However, there is an urgent need for the techniques to be adopted by anyone using the Web for presentations.
This will result in changing
to

There are a large number of browsers currently available. Netscape is the most commonly used at present. The following discussion refers to version 1.1N. The next version, in beta at the time of writing has a different method.
In order to give a reasonable size for text, the Preferences menu can be used to set the font size to Huge. This gives 24 point normal font. However, it blows up the Header-One font to 48 point, which is really too large. To fix this requires more arcane skills.
Many aspects of X applications are controlled by resources
that are set in a resource file. The major file for Netscape is
/usr/lib/X11/app-defaults/Netscape (or wherever your
system administrator has placed it). The Netscape documentation rightly
warns against using or messing with this file. However, it is not too
bad: edit it by replacing the line
*documentFonts*variable*huge.6.pointsize: 480by
*documentFonts*variable*huge.6.pointsize: 240Similar changes may be needed to other fonts or sizes depending on your local X configuration.
Netscape version 2.0 uses the General Preferences - Fonts menu to set font sizes. Choosing a size of 24.0 results in reasonable sizes for the headers.
So, depending on your browser, and the environment it exists in (Unix, Windows95, etc), the fonts may be easy or hard to adjust.
.Xdefaults file of
Mwm.netscape.clientDecoration:(i.e. to no value.)
This will improve the display to
How much should be put on a slide? This is the same as any other slide presentation guideline: as little as possible. A title should usually be there, to give context. A few bullet points should be there. That is probably about it. No more than seven items is the general rule. Of course, an image or table may be used to replace text items, but again there should not be so many of these that they give a cluttered presentation.
This would generally ensure that the items of one page all fit on one screen. Should this rule ever be broken? I have found that there are some instances where a long piece of text is difficult to split across multiple pages (such as computer programs). In such cases it may be better to fit more on one page than can appear on one screen. The problem of only seeing a part of the total text may be avoided by printing the page for a handout in such as way that all of the text can be seen on the one printed page (this is addressed again later).
An entire paper, that is scrolled through, is really bad news.
These buttons use up space on the slide. Where can they be put to minimize wastage? I generally put them either side of the title of the document. There is usually room there and they do not detract too much from the title (in fact, they help to highlight the title).
This looks like

Systems designed for presentation such as PowerPoint use a mouse click to move to the next slide. Such a capability does not appear possible in the current generation of browsers because they are designed to handle non-linear flows. The Next button will have to be explicitly selected wherever it is, which is a bit of a nuisance. It is also a bit of a nuisance that the buttons always appear right next to the title, which means they move from slide to slide according to the width of the title. Being able to fix them on the edges would be more consistent.
The default colours for Netscape are black on light-grey. This is a good choice for personal work. For presentations, there needs to be a much higher contrast. Yellow text on dark-blue is a common choice.
This looks like

Again, the ease of setting up such colour schemes depends on the environment and upon the browser.
With a collection of, say, twenty of more slides, there will be at least twenty Next links to be set up and maintained. Adding a new slide in the middle requires changing the link in the previous slide as well as setting up a link in the new slide. If there are also Previous links then changes have to be made in the following slide as well. This is tedious, and does not fit well with other slide presentation mechanisms that allow use of views to manipulate the entire structure.
Unless your document preparation system supports such links, an external
mechanism is needed to automatically maintain the links.
I currently use the Linux
documentation system in which the slides are written in SGML using its
own DTD. Tools that are part of the Linux documentation system then
automatically generate the slides with links between them. Changing the
order, or inserting another slide is done by a rerun of
linuxdoc2html on the original single SGML document.
The original Linux documentation system has a richer set of links than are required for a linear presentation, and do not use space-saving techniques. I modified parts of the system to produce a slide format that I was happy with. This required changing some C code and also changing some of the SGML to HTML translation tables. This is too complex for non-programmers, so a system that is easier to customise may be preferable.
The requirements of such a support system are reasonably simple
The use of an intermediate language such as JavaScript may help in this.
A student project at the University of Canberra is to build a browser in Java with the possibility of such special-effect mechanisms.
I keep a copy of Netscape 1.2 just for printing, as this allows control over the print characteristics - this has been removed from Netscape 2.0.
There are even worse problems with printing, such as how the browser deals with newer features of HTML such as Forms, Tables and Frames. All current browsers fail to print the contents of Forms correctly, and often fail on the other features. I use Forms with TextArea's regularly for interactive demonstrations: in order to print these I have to resort to even more obscure methods: I have a copy of Mosaic since that is one of the few with source code availability. This has been modified to print the content of TextArea's - again, a technical task. The future in this will become even worse with the advent of dynamic Java programs running in the browser!
Once printed, how is this collated with the rest of the slides? Using a system with multiple slides generated leads to problems here. Lacking a system to print a set of pages by following links, you have to print each page separately - which can be very tedious in a 100 slide presentation!
The separate pages then have to be reduced and reprinted.
I ``print'' each to a Postscript file, concatenate them into one,
and then use a Postscript
tool such as psnup to reduce each page and fit four pages
into one, generating a new Postscript file as a result.
Although this process is quite easy (once set up), it is very time consuming and needs to be improved upon.
The problem areas include
| Pointers to other Papers | |||
|---|---|---|---|
| Conference Presentation | Papers & posters in this theme | All Papers & posters | AusWeb96 Home Page |
AusWeb96 The Second Australian WorldWideWeb Conference ausweb96@scu.edu.au