PDF is a quite crappy format. It was never intended to be used for documentation. Intention of PDF was only! for producing output to be sent to a printing press. That’s why it’s divided into pages, and does even know what a “sentence” is. It’s just letters and words smeared over a page. You’ll probably notice this when you try to select text on a page where it’s printed over multiple columns. It’s a bit unfoturtunate that about 30 years ago datasheets transferred from being printed in paper books to being distributed as PDF files. First via CD-Roms, and later ever more directly over the internet.
Lately I’ve started experimenting with Asciidoc. It is a very simple “flat” text based format, and it can be “compiled” into HTML, PDF, epub and several other formats. It is a very lightweight format with very little overhead. It’s also the native format in which KiCad’s documentation is maintained. The text editor I use (Geany) has a few asciidoc extra’s, it for example shows text in bold and some color highlighting. I’ve also installed a plugin in firefox so it can load asciidoc files without having to “compile” them into html first. I just save modifications in my text editor, and about a second later, the modifications show up in firefox. This all works quite nicely.
I have become quite allergic to “word processor” formats. I once made a 70 page word document with a bunch of graphs and tables, and during work on that file microsoft word crashed a bunch of times. I even lost some work because I had to go back a few days in my backup to get a file that did not crash word. I do have LibeOffice installed on my Linux box these days, but I never used it in a serious manner. It feels to “convoluted” to me. I want a format I can understand, and in which I can be confident it will always be readable and easily maintainable / modifiable. And Asciidoc gives me that confidence.