subbu.org

JSON vs XML wars are still going on, and they are becoming more amusing, I must say. The latest is from Sean McGrath arguing about the significance of mixed content in XML.

Sean argues and that mixed content is at the heart of the difference between XML and most other data representation languages. I think that’s an extreme position. Mixed content is important in certain applications, but saying that mixed content is at the heart of the debate is a definite over-statement, and completely misses the point.

In fact, mixed content is one of the most avoided features of XML when used to exchange data. Processing mixed content is hard, and most developers avoid it. Some of the popular data binding APIs avoid it too. In data centric use of XML, mixed content can easily be avoided.

That is not to say that mixed content is bad. It has its applications. Would I use in data-centric applications? Most certainly not.

As I have argued before, the debate between XML and JSON is over. The key point is ease of use. I bet the XML pundits can come up with a zillion reasons why XML is better - but that is besides the point. The question is not whether XML is better over JSON as a general-purpose data interchange format. The question is the ease of processing data in a browser using JavaScript, i.e. the number of lines of JavaScript needed to process the data, and JSON is the clear winner. All other points made about XML vs JSON, are, IMO, pure distractions.