Am 10.06.2017 um 13:17 schrieb Junio C Hamano: > Andreas Heiduk <asheiduk@xxxxxxxxx> writes: > >> When setting `.gitattributes` in a second worktree, a plain `rm .git/index` >> does not actually delete the index. >> [...] > Right. > > I however have to wonder if we can do the same without futzing > directly with the "index" file as a filesystem entity. With or > without your update, what is taught in the document feels like > munging a disk block with binary editor to correct a corrupted > filesystem X-<. IMO `rm .git/index` is like munging a disk block WITHOUT a binary editor but with plain `dd seek=... skip=... count=...`, `hexdump`, `ed` and back - every step is clear in principle but painful and dangerous. :-) > For example, can we do this "empty the index" step with things like > > $ git rm --cached . That would be `git rm --cached -rq .`? Executing this in the git repo gives me an index file with 2.1kb. I don't know whether or not this index still contains something relevant for this case. > or > > $ git read-tree --empty > > instead? Nice! The `index` file contains 46 bytes. For me THAT one is like a nice binary editor apt for the job :-) I'll queue that.