Dydd Gwener 5 Ebrill 2002

Technologies

Hand Coding Versus WYSIWYG

When Tim Berners-Lee invented HTML, as a computer language for sharing documents between otherwise incompatible computer operating systems, he didn't actually envisage anyone writing HTML. He thought the translation of pre-existing documents would be handled by software. However, he was largely thinking about physics technical papers, which break down easily into text and headings. Most web pages now are written for the screen and don't translate well into print - and they do not share a common structure.

Berners-Lee didn't foresee the popularity of the web, or how quickly the technology would develop. Early documents were plain text. It was Mark Andressen, then a graduate student at the National Center for Super Computer Applications (NCSA), who added the capability to display graphics in a web browser. The NCSA browser he developed was called Mosaic, and both Internet Explorer and Netscape Navigator are based on it.

The conflict between the main browsers led to a slew of proprietal technologies, which rather than helping transfer files across the web, meant many would only display as intended in the one or the other. While the current generation of browsers (Internet Explorer 6, Netcape 6, Opera 5, Konqueror) display documents in much the same way, not everyone uses them. The only solution is to know HTML and not rely on quickly outdated authoring programs.

We use Arachnophilia, a glorified text editor, with quite a few nifty shortcuts for editing HTML. This means we don't get trapped into thinking that whatever my current operating system is showing me is anything like what anyone else gets. WYSIWYG is fine for editing print documents; for the web, it's misguided.

Server Side Stuff

The really important side is the server side. That's what happens before the page reaches the browser, that's where the shopping cart is, where your databases are, and where you can detect what kind of browser is being used, so you can send out appropriate DHTML. For all of this we use PHP. This means that all my pages actually break down into smaller files assembled into a page like the one you're reading now. Changes to navigation are easy - an extra link can be added to every page in minutes, no matter how large the site. PHP has several database functions too, so the information behind the site can be as complex as you like.

Got Validation?

We take great pains and great pride in making my pages valid with the W3C and their style sheet checker. This means that pages should work, not only on media I've haven't tested them on, but should work on ones not released yet. For more information on Web Standards, try WaSP site. Finally, my sites are designed to comply with requirements of Bobby.

af'ter-image
an image that persists for a time after one looks at an object Chambers English Dictionary
© 2001 afterimages.co.uk
Contact: David Weeden dave@afterimages.co.uk
Phone: 029 20 634 988