As far as I’m concerned, any sort of UI framework is a godsend. When first learning to design a visual UI structure, the first thing that I (and most people probably) thought is that it we’re going to have to write a mountain of buggy code just to make the most barebones interface imaginable. Although this mountain in some manner actually exists, UI frameworks essentially filter it down to the most essential and important aspects in an incredibly readable form. My first impression of this madness was the fact that what I’m creating will somehow have to adapt itself to the users window and preferences. That is, if they have a really large screen, or if they like to have multiple windows open… imagine going through all of the painful math and testing just to make your little page shift itself.
In this day and age, a firm understanding of these frameworks is essential to any sort of UI interface regardless of whether or not you’re hooked to the net. Just this summer I was working on some add-ins for Microsoft Office products, and guess what- the actual interface and window design is entirely in CSS and XML. These are programs that any educational institution requires you to have.
In retrospect, UI might actually be the most technically demanding stuff a computer would have to go through short of running virtual simulations of atoms and whatnot. Many of us computer science folks like to talk about how buff and customized our computers are (even Mac owners to some extent). We like to talk about our framerates and how we overclocked our processors and how our graphics cards are top of the line… but the people who require the most powerful tech are actually the people who develop the stuff we play with, not the kids who play. Do you know how technically demanding it was to produce Avatar (2012)?
Personally, I have no complaints. With all of this stuff in mind, semantic UI is actually kind of fun. Just stick an extra line here and there and suddenly your code jumps off of the page.