I posted this as a response to another post, but thought it deserved it's own discussion.
One of the difficulties with using jQuery Mobile is that it fixes a fundamental disconnect between the design of HTML and the real world we live in. HTML doesn't really model the latter very well. Maybe the designers were thinking ahead, and that it's not necessary to model the way we write on paper - e.g. "let's move beyond paper".
I sometimes get surprised reactions when I tell people that there are no pages in HTML, and so jQuery Mobile retrofits-in pages. So, of course there are some problems trying to make HTML and browsers do something they were never intended to do.
When we think of "pages", the physical implementation it most commonly brings to mind is a loose sheet of paper, typically of a uniform and standard size, or a leaf of a codex, or bound book.
The HTML specifications never mention pages. Only documents. And HTML only deals with one very specific type of document. Here's a physical analogy of the kind of documents that HTML deals with:
Which is kind of funny, since this isn't really very common historically. Horizontal scrolls, with "pages" or at least columns, are more common historically:
So, HTML is about 2000 years behind the times. (codices are first mentioned in the first century...)