Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

all: should not need to have separate key for ann+snapshot@example.com #194

Closed
robpike opened this issue Feb 20, 2017 · 5 comments
Closed
Assignees

Comments

@robpike
Copy link
Contributor

robpike commented Feb 20, 2017

Snapshots should be easy to set up, and one way to help is not to require the user's key to be stored twice, a chance for mistakes and skew.

Can KeyServer.Lookup ignore the +snapshot part, at least if a first lookup with it fails or some such?

@robpike robpike added this to the release milestone Feb 20, 2017
@edpin edpin self-assigned this Feb 20, 2017
@edpin
Copy link
Collaborator

edpin commented Feb 20, 2017

Agreed. We already give +snapshot special treatment in the DirServer, might as well avoid the key hassle. I'll handle this on the key/server side.

@adg
Copy link
Collaborator

adg commented Feb 20, 2017 via email

@robpike robpike removed this from the release milestone Feb 20, 2017
@edpin
Copy link
Collaborator

edpin commented Feb 22, 2017

@n2vi for final security consideration.

However, I just realized this wouldn't help anything. For two reasons:

  1. Even if the keys diverge, the non-suffixed user will continue to be able to see the snapshots dir entries (metadata), as per the special dirserver treatment.

  2. The only time it would be a problem is if the +snapshot user is trying to access files as the +snapshot user in the +snapshot root and discovers its keys no longer work. Then, even if the keyserver sent the owner's public key to it, its own private key wouldn't decrypt the contents.

What I'm getting to is that implicitly sending the public keys of the owner in response to the +snapshot look up would not help when keys diverge. And it would seem to one paranoid about security that the keyserver is lying.

For these reasons, I now think we shouldn't.

@n2vi
Copy link
Contributor

n2vi commented Feb 23, 2017

At minimum, we need to display a warning to the user at key rotation. If they're doing it because they believe their old key is weak or compromised, they need to think hard about snapshots.

@edpin
Copy link
Collaborator

edpin commented Mar 2, 2017

I think this is mostly a duplicate of #256 which was closed in favor of #269.

Tentatively closing this one in favor of #269.

@edpin edpin closed this as completed Mar 2, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants