Michael Eischer

The general rule is that commands that modify the repository will ignore `--no-lock`. Most other commands should support the flag. I wonder whether restic should just return an error if...

There are three options to advance this issue: 1. either we just document which commands support `--no-lock` somewhere. However, that leaves the problem how users would know where to look...

> * Testing this change requires mocking the `Repository` struct in `checker`. It looks like a lot of work to test a mock… WDYT? Is there a better way? There's...

> This still yields a hash mismatch (different file, it is different each time): If the hash mismatch keeps changing all the time, this likely indicates data corruption in the...

> Update: I am trying to increase my Vcore now with small steps. Increasing it has as affect that mprime fails less often One the one system on which I've...

> But, we need to also define the cases where the recovery is not successful. Those don't really exist for the backup command. If some data fails to upload, then...

Related to https://github.com/restic/restic/issues/1620 . The main problem with adding file hashes is that these likely have a significant performance hit. In particular, common hashes like sha256 are among the bottlenecks...

`restic snapshots --json` is also an option that shows multiple snapshots at once.

I guess we could introduce a counterpart to the `--compact` option. Although this can get messy rather easily as there might be interest in showing lots of different columns. The...

> It just seems like the most useful default the size of the snapshot and the size of the repo... What do you mean with "size of the repo"? That's...