Hi all,
I’m finding myself using convoluted, old-fashioned solutions because, AFAIK, I can’t trust Git for diffs of .md, Hebrew directories. I’m sorry, this is a little tedious post, but maybe, even if there is no solution, the workflow may interest you. If you have a better one - I’m very interested to hear about it!
Context: I’m working with ChatGPT. Every morning, I expand my Zip of .md files to a sub-directory of Obisdian. I edit these files using Obsidian, and both ChatGPT and I manually create new .md documents in another directory (WIP - Work in Progress).
Then, in the evening, I ask ChatGPT to generate a new Zip containing the changes. I expand the new Zip to a directory, and compare the evening Zip to the morning one to verify that ChatGPT did not take excessive freedom with my documents.
Both Zips are directory trees, and I’m using Beyond Compare 5 (BC5) to compare them.
This works reasonably well except for several issues:
First, BC5 “diff” for Hebrew isn’t the greatest, and I have to read the text left to right.
it of feeling the you give would This
Second, BC5’s display of .md files is text-only and unformatted.
Third - This workflow is basically “version control gone manual!” This can’t be the correct way to work!
Any ideas?