I have a repo called WebScraper with a single branch called main. I am the sole contributor to the repo but do so from multiple machines. There is NO gitignore file. The program scrapes web data and updates a fairly small csv (<2k lines). It runs daily on machine1.
Currently on machine1 CSV was edited this morning (1/25) and I can verify there's new data as the last lines are dated 1/25. Currently on web repo CSV was updated on 1/23 and I can verify this as the last lines are dated 1/23.
Inside the WebScraper dir I use git push and get a 'Everything up-to-date' message. A. Why does it fail to recognize that CSV has been recently updated?
git push
Next I try deleting CSV from the web repo and retrying 'git push' on machine1. Machine1 commits some changes but the old 1/23 CSV appears on the web repo (verified via data inside).
B. Where is it getting the old CSV files from? How does it conjure a file into existence and then decide not to use the more recently modified version?
Next I try 'git push --force' from machine1. It goes through the enumerating, counting, delta compression, writing steps, says 100% done, gives comit numbers andsays --> main (forced update) BUT the CSV is still outdated.
git push --force
question from:
https://stackoverflow.com/questions/65893424/git-force-push-restoring-old-version-to-repo-instead-of-new 与恶龙缠斗过久,自身亦成为恶龙;凝视深渊过久,深渊将回以凝视…