Is it really possible to love jQuery? It certainly seems like it is from the numerous blog and forum posts I've read while trying to figure out how to make it do some fairly simple things. Many of the posts end with rather disturbing terms of endearment: "...this is why I just love jQuery" being a typical example. Yet I'm still not sure that our first blind date will result in a lasting relationship.
Perhaps if you spend your life building websites that incorporate the now mandatory level of flashy UI, animations, and interactivity jQuery is pretty much a given. At least it means that paranoid people like me who have Java, ActiveX, and Flash disabled in their browser actually get to see something. I got fed up with the sites you used to see (or, in my case, not see) a while ago that were basically just a large Flash animation - invariably with the focus on appearance rather than containing any useful content. But disabling script is generally not an option these days.
Mind you, I'm now finding the same problem with sites that are just a single large Silverlight control; though - being a 'Softie - I guess I do tend to trust Silverlight rather more than other animation technologies. Well, marginally more - I'm still a paranoid neurotic. You know what they say: "Just because you're paranoid doesn't mean they aren't watching you."
There's plenty of API information, and plenty of blogs that provide just enough to not quite get things working. As an example, I'm using the load method to reload a partial section of a page into a div element, and I want to change the mouse pointer to a "wait" cursor and display an indeterminate "Please wait..." image while it loads. The docs say I can flip the cursor for the page using the css method of the element that holds the partial page content (though they don't mention that it doesn't work on hyperlinks within it). And that I can display my hidden div containing the loading image using the element's show method.
But, of course, figuring out this is the easy part. Simply calling the methods one after the other to change the mouse pointer, show the image, load the new content, hide the image, and change the mouse pointer back again doesn't work. Instead, you have to chain the method calls so that they only execute after the previous one has completed - mainly, of course, because the load method is asynchronous. The "getting started" docs hint at all this without actually using the dreaded "a" word; while the "real programmer" docs are full of barely comprehensible tips such as "CSS key changes are not executed within the queue."
Of course, these days, asynchronous programming is a common scenario, so I'm a bit surprised that the docs don't just bite the bullet and use the "a" word from the start. But I guess there's another issue as well: no programmer with any remaining shred of pride would use separate callback functions. You wouldn't dare let anyone see your code if it didn't use lambda expressions for callbacks - even if you are still a bit frightened by them. Let's face it, finding syntax errors and debugging statements that cover twenty lines and end with a dozen closing curly and round brackets is not a procedure designed to aid mental stability or promote a restful programming experience. Especially when the typical error message is just the amazingly useless phrase "Object expected". So maybe the documentation people want to avoid using the "l" word as well...