From our initial question of was there a better way to manage a Web site, our exploration of XML and Cocoon delivered a resounding "yes." XML not only alleviated some immediate problems, but more importantly had a positive impact on fundamental workflow issues.
But we also realized that conversion of a Web site does not happen without some pain. The learning curve was probably the biggest impact we experienced. For our initial developers who were investigating XML and Cocoon for the first time with limited external resources and no in-house knowledge, we measured a three-month learning curve before they became fully productive.
Much of XML is similar to HTML, but much is different. The most significant difference is that XML is not "page-based" like HTML, so it requires a different conceptual understanding and approach. Whereas in HTML, you are basically creating pages, and the page is both your source and output; in XML you are structuring your content and designing different deliveries. It is a radically different way of thinking about your Web site. While some knowledge and skills are transferable from HTML to XML, the same knowledge and skills can also be detrimental to thinking within an XML framework. Even experienced HTML developers and programmers need to learn anew as they move into an XML-based architecture.
On the positive side, our three-month learning curve has sharply declined within the past year as external resources (books, Web sites, user groups) have increased and our in-house knowledge base has grown. Experienced HTML Web developers brought on board since our Gateways Guide prototype has been completed now experience a learning curve of three weeks (versus the original team's three months).
| Next >