Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Werner Koch via Gnupg-users
On Thu, 10 Sep 2020 10:34, Martin Pätzold said: > the keys, therefore we had to extend the permissions for the > "private-keys-v1.d" directory to group access. I see. Just a hint: You may use the remote socket feature to run gpg-agent under a different account. It might take a bit of effort to

Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Martin Pätzold
>>> Long shot: does your system support ACLs? >> >> Using ACL would be possible, but we are reluctant to do so, since it >> adds a second permissions layer that is only visible if you actively >> look for it. > > Perhaps I am not understanding this correctly, but wouldn't that be a > good thing?

Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Jerry
On Thu, 10 Sep 2020 11:13:34 +0200, Martin Pätzold stated: > >> Yes, we have some period tasks that are handled by Celery. Celery > >> has its own user on the system and this user needs at least read > >> access to the keys, therefore we had to extend the permissions for > >> the

Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Martin Pätzold
>> Yes, we have some period tasks that are handled by Celery. Celery has >> its own user on the system and this user needs at least read access to >> the keys, therefore we had to extend the permissions for the >> "private-keys-v1.d" directory to group access. > > Long shot: does your system

Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Andrew Gallagher
On 10/09/2020 09:34, Martin Pätzold wrote: > Yes, we have some period tasks that are handled by Celery. Celery has > its own user on the system and this user needs at least read access to > the keys, therefore we had to extend the permissions for the > "private-keys-v1.d" directory to group

Re: private-keys-v1.d and preserve-permissions

2020-09-10 Thread Martin Pätzold
Thanks for the clarification and the patch. > Is there a special reason that you need to give group access to those > files? Yes, we have some period tasks that are handled by Celery. Celery has its own user on the system and this user needs at least read access to the keys, therefore we had

Re: private-keys-v1.d and preserve-permissions

2020-09-09 Thread Werner Koch via Gnupg-users
On Wed, 9 Sep 2020 19:37, Werner Koch said: > I looked at the history and the reason for the described behaviour is > documented at https://dev.gnupg.org/T2312. I re-opened that bug. Fixed in master and 2.2 see the ticket above for the patch. Salam-Shalom, Werner -- Die Gedanken sind

Re: private-keys-v1.d and preserve-permissions

2020-09-09 Thread Werner Koch via Gnupg-users
Hi, I looked at the history and the reason for the described behaviour is documented at https://dev.gnupg.org/T2312. I re-opened that bug. Shalom-Salam, Werner -- Die Gedanken sind frei. Ausnahmen regelt ein Bundesgesetz. signature.asc Description: PGP signature

Re: private-keys-v1.d and preserve-permissions

2020-09-09 Thread Werner Koch via Gnupg-users
On Wed, 9 Sep 2020 15:22, Martin Pätzold said: > And if the setting is not what I need, how can I prevent the > permissions for "private-keys-v1.d" from changing? The --preserve-permissions is a gpg option and not one of gpg-agent. In fact gpg does not known anything about private-keys-v1.d.

private-keys-v1.d and preserve-permissions

2020-09-09 Thread Martin Pätzold
Hello, I am working with Debian Stretch (9.13) and GPG 2.1.18. The "private-keys-v1.d" directory has per default the permissions 700 (drwx--), but I need them to be 770 (drwxrwx---). I can change the permissions ($ chmod 770 private-keys-v1.d) but after some time they are be back to 700.