Open securelyfitz opened 3 years ago
I'll add that I'm pro flat README :grin: I think the Table of Contents works great.
Just to add an extra opinion.
I found it relatively simple to convert the single README to an ebook as a little reference for on my e-Reader, only needed minor post-processing to fix the code blocks and add a ToC using Calibre, the "#" headers become chapters and sections etc.
Relative image URLs would help a bit too but not a big deal.
Love the flat README - as is it's pretty brief on each topic I think, so still hitting the README feel. If you start adding multiple examples to each section could see what you mean about breaking it out.
If you want a bit of both, you can setup github actions that do fancier processing on doc commits. For example with https://github.com/newaetech/chipwhisperer-target-cw308t we have a GH action that copies all the README.md
files in the individual subdirectory into another repo with different names & then builds the doc output with mkdocs to https://rtfm.newae.com . You could probably break out individual .md
files if you wanted but still auto-merge them into one flat .md
file for those who want a single file. But it's a lot more hassle than you've got now ;-)
I would love to have a downloadable, self-contained PDF or EPUB that can be viewed offline.
I would love to have a downloadable, self-contained PDF or EPUB that can be viewed offline.
Pandoc is pretty good at converting markdown into a readable PDF/EPUB. I don't like the idea of keeping generated binary PDF/EPUB files in the repository, but perhaps this is something we could include in releases like we do with the schematic PDFs.
Agree with @fharding1, i like including a pdf/epub in the release bundles. We're overdue to release the v1.1 bundle, but i have a few readme more additions i should include before we roll that.
Previously considered refactoring the readme into separate readmes per interface, but have gotten lots of feedback on the 'flat' readme format, so will keep it.
I added a toc in 1e7971c34c387b02b1f95d27a5c4665b8327651b (first generated with doctoc, but manually tweaked for brevity - it should be rather static so no need to automate updates)
Originally the 'usage' was supposed to be broken down by header, but became a list of protocols instead. Then, the 'pinouts' was supposed to be a raw pinout list, but became a protocol specific list instead.