I write it directly in hand–coded HTML. I used to use Windows Notepad to do so, mostly for the hair it put on my chest, but now I tend to use program editors, particularly xCode. I don’t use fancy site building software, as you can probably tell.
The problem with such HTML pages, indeed any kind of computer code, is that computers only understand precise syntax, and it is very easy to get that syntax wrong. That’s why I’ve built a bespoke publishing process that checks for errors.
I don’t use a specialist web page editor, mostly because they’re not built for people like me. I find most too formulaic regarding site design. I’m quite sure the super whizzo professional products don’t make such mistakes, and I’m quite sure too that I can’t afford those products!
That’s why I edit this site’s HTML web pages by hand. The publishing process compiles my original HTML with Server Side Includes (SHTML) pages into pure HTML, although the SHTML extension is retained. Indeed, all dynamic pages, including archived .ASP and .PHP pages, are rendered into pure HTML.
I’ve written an accompanying suite of small support scripts for preparing photo pages for publication. They add this site’s standard exif values (foto.sh), and do most of the legwork for preparing photo page translations (cpfoto.sh). Those two scripts make rather naughty presumptions about where they’re run.