While HTML is known commonly as a "language", it isn't one. It's really just a set of formatting tags applied to plain typed text. Despite its simplicity relative to a true programming language, it can be a pain in the neck to use. Further, the only tools originally available to edit an HTML document were straight text editors. Not exactly a feature-rich category of software. Finally, when building a page using a text editor you don't actually see what it's going to look like until it gets posted on the Web.

Enter Web production products, whose reason for existence is to eliminate the above problems, and then some. You can see dozens of them on the market. Microsoft Front Page, Netscape Gold and Adobe PageMill are a few of the heavy hitters in this arena. These babies will darn near design your site for you, and are especially good for novices.

At The Webmaster we've tried many of these products. On the surface they look pretty slick. However, we've always gone back to a straight text editor to design web sites. There are several reasons for this:

We learned HTML back before these products existed, and are most comfortable working directly with a raw HTML document.

Without exception, the HTML produced by these fancy web production packages is bloated: the structure is inefficient compared to a document produced by hand. Big files translate to increased loading times for your viewers. The web is slow enough already. We won't produce a web site that isn't as fast-loading as possible.

When we make the documents ourselves, a "purer" strain of HTML comes out than that found in most production software packages. There's true HTML and there are "extensions" to it... and those extensions aren't always supported by all browsers. Some standard HTML tag shows up differently depending on your browser, despite the tags' supposed standardization. We use only the genuine article, and hand- (mouse?) tweak it to take into account variances that can be expected when users show up with different browsers.

So, is everything done by hand? No. HTML may be relatively simple as computer languages go, but nobody is perfect and typos happen. That's why we use the latest HTML "validation" routines to ensure the work that looks like it's just fine really conforms to HTML 3.2 standards. What is a validation routine? Think of it as a spell-checker for HTML syntax.