Why not Asciidoc? It works well with AI so far. To make it even work better I made a more formal subset which can be read with a single-pass parser (https://github.com/rochus-keller/leandoc/), to reduce the "level of surprise" for all parties. Why should an xml-kind format be better suited for AI?
1 comment
[ 2.6 ms ] story [ 13.2 ms ] thread