|
Everything Jellby said. And also, don't forget to think about the future. Loads of ebooks look fine now, but under the hood they're a ticking time-bomb that might explode when converted into some future format.
The main point of separating the semantic and presentational aspects of the code is that it makes it very easy to modify the presentation to suit any changing conditions. If a future format comes out that offers some compelling enhancements, which is better: changing a couple of lines in the css or having to do an exhaustive search and replace on the entire code?
It's pretty obvious that many people are having a hard time producing good-looking ebooks as it is, even when working from digital sources. The way to avoid this is to ensure that the code is structured as clearly as possible using the semantic/presentational distinction.
|