Use the source, Luke

FCW's DotGov Thursday column advises Web developers to go to the source learn and understand HTML

With the wonderful range of Web development tools available — from simple

WYSIWYG editors to multisite management applications — it is quite easy

to build rich, full-featured, sophisticated Web sites.

This is, for the most part, a good thing.

However, there is a downside to using tools like FrontPage, NetObjects

Fusion or Dreamweaver. It's not a problem with the tools themselves. The

problem is a lack of knowledge of HTML by many who use them to develop Web

sites.

Everyone who does any work building Web sites should have a fairly good

understanding of HTML specifications and Cascading Style Sheets.

Sure, today's Web development tools let you get away without knowing

HTML or CSS. Point, click, drag, drop, publish, and then it's Miller Time.

Nevertheless, I maintain that even if you never hand-code a Web page in

your life, you still need to understand the basic, fundamental foundation

of HTML in order to get the best out of the medium in which you are working.

When I started working on the Web in late 1994/early 1995, most of the

Web development tools were just text editors with a little functionality,

such as prefabricated tag generators, added on. Most people used vi or Notepad

to write HTML.

Now I'm not saying that everyone should go back to using text editors.

It would be impossible to do so and maintain still any fairly large site.

It is, however, helpful to know what to look for, and where to look,

when something doesn't look right ("Why is that table not aligned correctly?")

or when visitors to your site have problems ("The menu on your site doesn't

work!"), yet it all looks fine and works well for you. Understanding how

the code works will allow you to make better use of the WYSIWYG tools, too.

There is another problem where being able to understand the source code

for a Web page will come in very handy. Most of the WYSIWYG development

tools have a tendency to generate source that is not optimized and may add

lots of extra text that can bloat the size of your pages. This isn't something

most people who are writing the pages notice.

Generally, the Web developer has a local-area network connection to

the Web server, so pages just pop right up. However, most of the time your

site will be visited by people connected to the Internet via a much slower,

dial-up service. Extra text and even extra white space can slow a page down

enough so that users will just hit the stop button.

Everyone knows about keeping the graphics small to help speed up downloads,

but how many know about the performance issues that occur because of the

way HTML is rendered?

Understanding the source is, as Obi-Wan might have said, the key to

understanding the Web.

— Klemmer is a senior Unix system administrator and security analyst

at the Strategic and Advanced Computing Center at Army headquarters. He

can be reached at joe.klemmer@us.army.mil.