[Puppet Users] Re: Proposed refactoring of ssh_authorized_keys

2008-10-14 Thread Paul Lathrop
On Mon, Oct 13, 2008 at 6:16 AM, Ryan Steele <[EMAIL PROTECTED]> wrote: > Maybe I'm missing something here, but how is that even possible since > the namevar is not unique? Puppet would produce an error when trying to > declare that second resource, since it has the same name as the first. > If t

[Puppet Users] Re: Proposed refactoring of ssh_authorized_keys

2008-10-13 Thread Ryan Steele
Hey folks, Before responding to the ticket below, I think I should provide a little bit more qualifying information. I had a discussion with Luke Kanies regarding the design of the ssh_authorized_keys type, and the agreement we came to was that the key itself would be a good identifier becaus

[Puppet Users] Re: Proposed refactoring of ssh_authorized_keys

2008-10-12 Thread Peter Meier
Hi > You're correct, let's move that discussion into the mailing-list > instead of chatting in the tickets. +1 > In #1644 you wrote: > >> In my opinion, the proper design would be to have ONE >> ssh_authorized_key resource per user, and that you should be able >> to provide an array for both t

[Puppet Users] Re: Proposed refactoring of ssh_authorized_keys

2008-10-11 Thread Francois Deppierraz
Hi Ryan, You're correct, let's move that discussion into the mailing-list instead of chatting in the tickets. In #1644 you wrote: > In my opinion, the proper design would be to have ONE ssh_authorized_key > resource per user, and that you should be able to provide an array for both > the "tar