Skip to content
This repository has been archived by the owner on Apr 26, 2024. It is now read-only.

'Web of trust' for automatically trusting certain keys without human verification? #2712

Closed
andswitch opened this issue Nov 24, 2017 · 4 comments

Comments

@andswitch
Copy link

andswitch commented Nov 24, 2017

Hi,

Matrix is awesome, but the manual per-device key verification mechanism is a bit of a pita. Perhaps it is possible to make this mechanism a little bit more user-friendly by automatically trusting certain keys in case these keys have been verified already by a trusted party?

Especially in case a user A with a device X uses device X to verify that another device Y belongs to him/herself. Anybody who has verified that device X belongs to A could trust that device Y also belongs to A. This way, once you've verified one device from a user, you can trust all devices from that user as long as that user makes a little effort to verify those devices. Then, that user doesn't have to annoy other users with the need to verify the other devices.

@andswitch andswitch changed the title 'Web of trust' for automatically trusting keys without human verification? 'Web of trust' for automatically trusting certain keys without human verification? Nov 24, 2017
@uhoreg
Copy link
Member

uhoreg commented Nov 24, 2017

@erdii
Copy link

erdii commented Jan 12, 2018

sounds a bit like keybase. where each user has a ledger of his device keys and uses already trusted devices to cross-sign new keys

@ShadowJonathan
Copy link
Contributor

With the introduction of cross-signing in E2EE, could this issue be considered solved?

@uhoreg
Copy link
Member

uhoreg commented Sep 18, 2020

No, this issue seems to be more about a general web-of-trust mechanism, whereas cross-signing is specifically about same-user verification.

That said, this is more of a spec issue than a synapse issue, and we have https://github.com/matrix-org/matrix-doc/issues/1886 to track this on the spec side now (and element-hq/element-meta#662 as the user-visible feature issue), so this should probably be closed anyways.

@uhoreg uhoreg closed this as completed Sep 18, 2020
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants