Most browsers support jumping to specific links by typing keys defined on the web site. On Windows, you can press ALT + an access key; on Macintosh, you can press Control + an access key.
All pages in this document define the following access keys:
These files are written using HTML 4.01. HTML 4.01 Transitional is used, rather than XHTML, because that is what I started with and XHTML is harder to code by hand. I’ve not found any reason to change. HTML doesn’t have to mean horrible non-conforming code. Having some end tags optional, for example, is part of a formal specification, and the pages are validated against that specification.
Each page contains a HTML 4.01 validation logo in the footer. Clicking on the logo will revalidate the page using the W3C’s Markup Validation Service.
If Javascript is disabled or the page is being served locally, it will go to the Validation Service main page but not automatically enter the URL to be checked.
Browsers support navigational links that are coded in the page’s header via <LINK REL= …> statements. In addition
to the visible hyperlinks in the document presentation, these provide standard hooks for shortcut keys, assistive technologies, text-only browsers, and browser user interface widgets.
In the Firefox browser, you can use the Link Widgets or cmSiteNavigation Toolbar to make direct use of these. Note that they don’t show up if the page is in a frame. Firefox only uses these links if the page is loaded as the top-level document in the window.
See §6.12 Link types in the HTML 4.01 Specification.
Next, PrevSection, SubsectionLinks are typically written to make sence out of context. I eschew link text like “click here”. Usually a link will be the name of the thing being cross-referenced,
as naturally mentioned in the text. If the meaning of a link is not clear from the text enclosed in the A tags, then a TITLE attribute will be used.
All images include descriptive ALT attributes.
These pages use cascading stype sheets, and avoid mixing content with presentation. The text is marked up using proper semantic HTML tags, such as UL
for a list, with CSS then used to make it pretty.
The document will be readable with CSS disabled.
No absolute font sizes or column widths are used! Layout of text is fully “liquid” and will flow to fit your window size and your text size.
Frames are used to provide navigation panes around a content window. If frames are not available, or only one frame is (easily) used at a time, the document can be navigated through its own hyperlinks, “See Also” statements throughout the document, a “Related Documents” section at the top of each page, Navigation Links and Access Keys, and links back to the Class Summary and document main page.
The earliest design principle adopted was to make the entire documentation set downloadable and browsable locally. That means it cannot be generated dynamically by code on the server, but must exist as a set of static HTML documents.
Client-side scripting was originally not used at all, and as much functionality was provided as was practical. Then, Javascript was added to enhance the usability and ease of development. Javascript is not required to read this. It meerly enhances it. Specifically,