Greetings, Otto! I use simple bash script, that makes a list with all unique words (or, rather, entities) inside the whole source tree, and then I look through that list semi-manually, sometimes picking out words nearby. Running this script for two different states and diff-ing two files has decent result, although resulting list of unique words still needs to be looked through, as it contains a lot of entities which neither are really words or typos. (I've found a decent amount of typos made between 49a6baec and bb9f0104, for example "acces", "afile") This script in theory should also be able to find unique/deleted identifiers/functions/e.t.c., although I myself didn't find any yet. The script itself looks like that: find . -type f \! -path 'unwanted_path' \! -name 'unwanted_name' \ -exec grep -o -P '\b[a-zA-Z_][a-zA-Z0-9-_]+' {} \; | ... | sort | uniq -u Sorry I've kept you waiting. Best regards, Vasilii