Closed sshilovsky closed 2 years ago
Please verify that the test on PR #73 is not skipped when it shouldn't be (3a1d2a1e265c43971f78ed1845920539ed00fd35). It's the first time I see ruby.
Released as 1.16.1. Thanks!
(I left out the change for skipping the test if unshare failed, because I'm uncomfortable with silently skipping tests if something goes wrong. It'd be better to change the test system so that a single failure doesn't block other tests from running, but I have too little time to work on bindfs these days.)
Thank you for prompt merge!
On unshare
, it is a step in the process of initialization environment for that particular test. I think that it failing can only mean that the initialization is impossible here, so the test can be skipped, just like it's skipped currently based on parsing man unshare
.
Although there is another possibility that my chroot, which I was running the tests inside, was set up wrong. I should look into that anyway.
Yeah, making the tests work in a variety of environments is a totally valid way to look at it (and some test frameworks explicitly support it), I just tend to put a lot more weight on dangers of invisible failures.
--map-passwd-rev and --map-group-rev
My use case for --map-group-rev is mounting /dev into chroot tree with different UID/GID. The other one was added just for completeness