About This Website

This website has been developed (almost) entirely using tools and systems that I have developed over the years. No overall supporting content-management system (CMS), such as Wordpress, has been used. I opted for this approach because of a desire to create a website in which information technology is used to help manufacture the website, not to help hand craft it.

On this website, I wanted to have a substantial amount of material that could be created automatically, such as annotated bibliographies developed from direct queries to PubMed or from material maintained in a separate citation-management system.

For a while I tried to connect these automated processes with Wordpress or other CMSs, but I kept discovering that the design of those systems had been optimized for the hand-crafting of websites and actually got in the way of using content that had been "manufactured" outside the CMS system. This current website now has more than 100 individual pages that together contain well over half a million external links. Maintaining all of this through some kind of computer-assisted hand crafting was just not feasible — an IT-assisted manufacturing process was required.

When I first started creating websites, more than 25 years ago, there were no generally available tools to help with the process. You created a web site by sitting down with a word processor and writing HTML code. I quickly discovered that that could be profoundly tedious for dealing with trivial changes.

For example, in one of my early efforts I tried to create a bibliography accompanied with photos of the authors. For variety, I wanted each entry to alternate left and right. That is, the first entry would have the photo on the left, the text on the right. The second would switch, with the text on the left and the photo on the right. This was well before CSS had been developed, so the best way to do this was to put the contents into <TABLE> structures. But if I wanted to add a new item to an existing bibliography, every existing entry from that point on had to have its left-right layout reversed. Yow! What a hassle.

So, I developed software that would allow me to maintain the data in a tag-value data file that could be fed into custom code that acted like a mail-merge program, with HTML coming out the other end. The format of the HTML output was controlled by a separate driver file that contained HTML intermixed with <<tag-names>> that would be replaced with tag values, as the data file was read. That program let me "manufacture" TABLEs of bibliographic content very quickly.

With the advent of CSS, I could just modify the driver files to contain HTML with CSS markup and then put the formatting into the separate CSS file. This gave me even greater flexibility.

Also, at the time I was doing this, I was mainly developing the WWW.ESP.ORG website that was providing access to classic scientific literature to the world. Many of the early users of the site were from parts of the world where access to desktop computers was limited, and the computers that could be accessed were old and primitive. For that reason, I wanted to stay away from any client-side systems that would put a computing burden on the user's computer. Similarly, since I was working on my server as a one-man staff, I wanted to stay away from technology that I didn't understand completely or that would require significant effort to secure and maintain.

RJR Experience and Expertise