A word of caution: The following text is written today, December 2025. Microsoft has not ended the system of frames as such, but no longer facilitates the frames pages. The relevant buttons have disappeared from the stock of buttons to be used with the word processor Word for Windows. I have adapted the original text accordingly. I believe that it could still inspire others to take part in the development of digital publishing.

Paul Grätz, Amsterdam, December 2025.

§ 1 Infrastructure

The aim of this paragraph is to provide structure to the development process that allows your books based on frames pages to be read on a computer and the Internet.

What is needed for this type of book publishing is Microsoft Office software, or in any event the word processor Word for Windows. The processor allows you to convert Word files into HTML files, and permits the setting up of frames pages. Before you begin, however, it is advisable to create a folder such as Atelier, and inside Atelier create two folders, named PNG and JPG. Let us suppose that your favorite book is Moby Dick, which is a fine and famous book, and you want it to be read in your own publication style on your own computer, and possibly also on the Internet, if you feel that your publication style may add something new and valuable to the number of publications of the book that are already out there. You would then create a Moby Dick folder in the PNG folder for all the relevant graphical files at your disposal with a png extension. Do the same with a Moby Dick folder in the JPG folder for relevant picture files with a jpg extension. PNG files are used typically for graphics consisting of several layers, whereas JPG picture files, in view of the compromise between picture quality and file size, have become the preferred choice for page background content. In today’s computers, the surplus of processing power and RAM memory ‘under the hood’ has disqualified the validity of the argument for the ultra-compressing GIF format.

Besides the word processor, for making frames pages, what is required is a screen capture program, for cutting parts off a frames page, and a picture editor, for producing layered graphical files. The procedure is then to use a screen capture program for cutting a JPG file off a frames page, then use the JPG file as a layer in a PNG file, and finally save the PNG file as a JPG file, to function as a background for the text or navigation bar frame in the web page. Why not use a frames page instead of a web page? The reasons are economy and flexibility. A typical frames page may contain up to a hundred or so frames. Each frame is bound by a size limit. Cutting one JPG file off the page reduces this number to just one. Moreover, it will allow you to manipulate the JPG file in a picture editor, which is a design plus (see f.i. The Third World and Bouwkunst (Architecture). For our purposes, a typical web page uses just two or three frames, plus edge frames to leave them all in the center of the monitor screen.

Screen capture software and picture editors can be found on the Internet. I have always used CaptureWiz and Microsoft Digital Suite 2006, respectively. The last-mentioned program allows for intuitive use of transparent PNG files in one single scalable working space, which is perfect for handling several PNG files simultaneously (see f.i. the left ‘page’ of the Tutorial picture). The use of the also still available Microsoft Expression Web 4 program may also be considered in the event of a need for restoring pictures which have lost their original acuity after too many size changes.

A web page is also a frames page, which means that there are actually two backgrounds, one picturing the background of a particular frame in a frames page, and a capture of the result in the frames page serving as a background in a similar frame but different page, the web page. The background is a JPG picture, one usually in a series of pictures designed with a particular purpose in mind, such as f.i. the sequential accommodation of sections of typed text, or a background supporting the appearance of certain controls, gallery and catalogue items, and so on. For a book, a screen capture of a frames page may be of its text frame, its navigation frame, or of any combination of these frames. Make sure here that the picture dimensions that you start with are larger in size than those that determine the frame dimensions of a frame in a web page. A text frame picture created in a picture editor ought to be larger in size than the same picture loaded as a background for a text frame in a web page. Reduction in size will result in sharper, more accentuated pictures.

When loading the picture files in a web frame, you may be surprised to find that its size seems to be incorrect for the frame. There may be two reasons for this: the JPG picture file may be set for a wrong quantity of dpi (dots per inch), which renders the picture content incorrectly (the safest dpi for the Web is 96), or the size of the frame seems to be at odds with the graphics used in the frames. To find the right dimensions for the JPG picture again is easy: click inside the appropriate frame for the mouse pointer to appear, then right-click with the mouse, select ‘frame properties’, and toggle between ‘relative’ and ‘centimeters’ or ‘inches’, but without actually altering and changing the result. If the frame format property is ‘relative’ to begin with, just toggle between widths and heights enough to get the correct earlier measured-out dimensions in centimeters or inches. You may run up against something here for which there seems to be no solution. Sometimes the altering of just one dimension, width or height, will be available, as is the case with edge frames. Either way, you will be provided with the stored, correct dimensions for the picture or, preferably, the canvas, in the picture editor. Another size issue is brought about by frames’ borders in a frames page. Because borders are shared by frames, removing them from one frame alone is impossible without removing all borders. Apart from using frame borders perhaps esthetically, or if very small frames are used, to keep them visually separated, it is best to work with the option ‘no borders’.

Edge frames are what makes a web page different from a frames page. Edge frames are meant to secure visibility and readability of the book pages. To that end the horizontal edge frames, both top and bottom, should be set for relative height, and the vertical ones, both left and right, should be set for relative width. The result may differ from screen size to screen size, but the web page will always secure center space for itself in the browser window. However, when opening a new frames page, it is advisable to leave all edge frame sizes ‘relative’, because here the web page is built up from within the center by loading special purpose frames, with dimensions based on dimensions set for background pictures. Word’s software will have made sure that corresponding edge frame dimensions have been provided with their fixed sizes. For example, opening a web page with two horizontal edge frames will lock both frames to the top and bottom sides of the screen. And, again, opening with two vertical edge frames will lock both frames to the left- and right-hand sides of the screen; for the subsequent horizontal edge frames there will be only one option left: a change from relative to fixed-sized for their widths, with their heights remaining relative.

The shaft or ladder structure on the left-hand side of the picture provides a clue with respect to the page number. Based on the position of the marker in brown, it should be page 5. The Introduction took 4 pages, and all the other positions are still vacant. When the marker occupies position 13, it remains there, and the page number would be 8 (13 – 5). The next marker would be for page 9, and terminate at position 12. The page number for these two fixed markers combined would then be 25 (13 + 12). In the pictured book project, the total number of pages prepared for text would here amount to 13 + 12 + 11 +10 + 9 + 8 +7 + 6 + 5 + 4 + 3 + 2 + 1 = 91 pages.

The navigation bar frame displays a choice that can be made for a combination of footnotes if applicable (NOTES page) and endnotes (NOTES All). Having taken into account a reader’s desire to read the complete text through without being distracted by footnotes, at the bottom of a page or sometimes even stretching to more than one page, no hyperlinks have been foreseen in the text on the page displayed above. NOTES page means that the same text as rendered in the TEXT page will be provided with hyperlink signs only if applicable. Absent any note on a text page, the caption NOTES page in the navigation bar frame will have been given a different (greyish) colour. Hyperlinks on a notes page link back to the relevant text page and, possibly, also to a second note page for lengthier notes. NOTES All and PAGES All speak for themselves.

Another option tried once was to develop for a document a system of normal scrolling pages, one for text, and another one for notes. Hyperlinks were positioned in such a way, that a reader could move from text page to text page, and from text page to note page, and from any other note page again to any other text page, enabling a browsing of the document not based on the inevitable Table of Contents at the beginning of books. There are many more interesting features of electronic publishing. Consider the Tutorial framespage pictured above with two identical sets of text pages, one for the left-hand side and one for the right-hand side. Forget Sticky Notes; start enjoying a real browsing experience.

§ 2 Superstructure

The aim of this paragraph is to discuss some more issues arising from a desire to make your software tools do things that they were not originally designed to do.