Just splitting out a discussion arising in another thread to one of it's own.
In traditional print book there is no line or space between paragraphs and the only way to tell it's a new paragraph is the indented text on the first line of the new paragraph. They also use a single blank line to indicate a change of scene, and a new chapter is an obvious major change in the story as well: usually a new page, part way down the page, and sometimes a drop capital. However, today there are more options available, especially in electronic formats where the display software will often remove any indentation of the first line, thus a blank line is needed between every paragraph and other methods needed to indicate a change of scene. Also, some software will reduce multiple space commands and multiple new line commands to display a single space and a single line, so something extra is needed to ensure the writer's intent is clear. With some software this same problem applies to the display of a new chapter.
When the software provides for displaying a new chapter as a new page to display that part is easy, but it relies on the software recognising the the correct place to break the text or having each chapter as a separate file. A lot of software recognises header commands like the H1 and H2 type commands in HTML, but not all software does, or is set to do so, and you can end up with more than one chapter displayed on the page.
Above are the problems, below are some of the common answers, followed by what I do. Do you have anything else you use?
For scene and section dividers some people use a full or partial row of symbols like * - _ . and some people use the same symbols with a space between each one in the row, because some software won't display multiple items in a row and some software will see a certain number as code and display something else (in some cases the code means don't display what follows. Often people will have a blank line before and / or after the line of symbols. In some cases people use an image between sections. Some people use a section title to mark the change. Some people don't bother with a specific section / scene divider at all, and leave it up to the reader to work it out from the context.
For my stories I use the one format for both print and electronic display, and have a mix of actions to show my intent. I do this because I want my intent to be evident, despite any changes in the presentation by the display technology. In the original word processing file I use the traditional options of position on the page, different font size, different font type, colour, and forced page breaks (i.e. new page) to show chapters, sub-chapters, and sections. For a change of scene within a section I usually use a group of underscore lines ( ____ ) to create a part page horizontal line.
The result is the chapter heading shows a third of the way down the page in large, bold, red text on a new page; a sub-chapter heading is in large, bold, blue text in italics at the top of a new page; section headings are in standard size bold text. In all three cases they're centred in the line with a space between them and the text below. I use the same formatting for the print books, the e-pubs, the html, and the tagged text code for SoL.
The only difference between versions I make with the code is the Partition command for the SoL code - this is the command SoL uses to start a new page for display to the reader on their site. I will place this at the start of a new chapter or sub-chapter because I try to provide a reasonable read of around 8,000 words with each chapter and SoL has a limit where it automatically splits the page after a set amount of characters which comes out to around 12,000 words, but also includes punctuation marks in the count. In order to get suitable breaks to match my headings I sometimes have a SoL page of 4,000 words and sometimes as high as 11,000 words. Despite trying to avoid a forced page cut, I sometimes end up with a multi-page display for some SoL chapters.
Some other display variations that occur are the fact the print books are in black ink on white pages, thus the colour vanishes. However, the position on page, the page breaks, the font sizes, italics, and bold all stay to convey my intent to the reader. With the electronic displays I get a variety of results. If the story is displayed as a single html file it shows as one long page and sometimes the reader's setting will adjust how it displays, but the H1, H2, and H3 commands usually come through and show the chapter, sub-chapter, and section headings as sufficiently different to do the job. The big oddity is the same e-pub file doesn't display the same way in all the e-pub reading software. I've seen the font size code removed, the font colour code removed, and the position on the page is removed in most software. Luckily, the forced page breaks, centre, bold, and italics commands come through in all the e-pub software I've seen. Thus there is a still enough variation between the various headings to carry my intent to the reader. What it means is: some readers get a more colourful text to read, along with more obvious break clues. However, this does highlight the need to be aware that not all display software provides the same answer from the same set of code in the file.
Does anyone do anything different to identify chapters, sections, or scenes? If you do, why?