An example for how to use RequireJS needs to be as clear and simple as possible, and putting a SCRIPT tag into HEAD is still the clearest and simplest way - the stuff that Souders writes about is Advanced JS hackery - if you understand Souders, then you also understand how to rewrite the RequireJS example to be even more efficient than it is.
One of the bigger misconceptions about RequireJS is that it's about efficiency. RequireJS does provide an efficient way to load up the scripts, but that's an important side benefit of the actual reason why it exists: To provide sane modularization and dependency management of your Javascript code.
thanx for the overview. Sounds like this falls into the same groups as LABJS and ControlJS to some degree.
I don't really agree that putting the script tag in the head is a simple example. The tornado web docs, talk about loading script tags last, and they cover a lot more than just where is the best place to put script tag.
I could buy the "it's okay to write demos with script tags in the head" bit in the same why I buy it's "okay" to use document.write() examples b/c it's 'fast'. Which is to say, I don't buy either argument.
If requireJS is aiming at people who don't know any advanced JS 'hackery', why on earth would that same dev know they need "sane modularization and dependency management". They still think it's "okay" to load script tags from the HEAD.
One of the bigger misconceptions about RequireJS is that it's about efficiency. RequireJS does provide an efficient way to load up the scripts, but that's an important side benefit of the actual reason why it exists: To provide sane modularization and dependency management of your Javascript code.