Simon Willison’s Weblog

Curious Javascript in .NET

26th April 2004

I’ve never had the opportunity nor the inclination to do anything with .NET; at work we use open source tools for all of our web development, and I prefer open source tools for my own personal experiments as well. At any rate, the javascript:__doPostBack links I’ve seen on .NET powered sites such as Channel 9 and Orkut plain give me the willies.

Anyway, I decided to view source and see what __doPostBack actually does. Here’s the function:

function __doPostBack(eventTarget, eventArgument) {
  var theform;
  if (window.navigator.appName.toLowerCase().indexOf("microsoft") > -1) {
    theform = document._ctl0;
  }
  else {
    theform = document.forms["_ctl0"];
  }
  theform.__EVENTTARGET.value = eventTarget.split("$").join(":");
  theform.__EVENTARGUMENT.value = eventArgument;
  theform.submit();
}

Note the use of a dubious user-agent based browser detection method. Now what I just don’t understand is why that browser detection is there in the first place. There are numerous ways of accessing a form by name using Javascript. document.forms['form-name'] is part of DOM level 0, and is supported by virtually every browser since Netscape 2. document.form-name is a Microsoft invention. The thing is, document.forms['form-name'] is supported by MS browsers as well. So why do a browser detect and use the MS only method when just using the other method with no detect would work equally as well?