Florian Wilkens

Hey good catch, seems I messed up the indices to cut out the wrong char :) The env support is not fully up to my standards yet anyway, but I'll...

That is indeed a valid use-case, that I did not explicitly consider before. I am currently thinking about how to implement that, as I want the `Client` to be as...

I think this should actually be `conversation_mut`? As you spotted the error, you can take credit and fix that or I'll do that later. Thanks for spotting the issue! :)

Sorry for the late reply! I was figuring out some details in `pam-sys` and `pam` was left in a weird state. I pushed some changes to `pam-sys` that hopefully improve...

So I should have looked at this PR first maybe ;) I was not familiar with the XSSO specification, as I modeled everything after the Linux implementation of PAM. Am...

I added some comments to [pam-sys/#16 ](https://github.com/1wilkens/pam-sys/pull/16), would be great if you could check them out :) Regarding this PR, I checked `security/pam_appl.h` on Linux and it already includes `security/_pam_compat.h`...

I merged your other PR first, so this needs a rebase, but after that I'll merge this. If you could also write a CHANGELOG entry, that would be awesome, if...

That's interesting which test did you run? I tried the `spawn_bash` example but on my machine I never got any valid pointers on the second level even for the first...

Could you rebase this once more on master and I'll merge it. Sorry for the delay :)

Sorry for the delay. It seems the commit history in this PR is messed up after the merges (duplicate commits). I tried to squash some commits but failed (at least...