Re: [openpgp] To bind or not to bind

[Kai Engert <[e-mail]>]

On 22.03.24 21:17, Daniel Huigens wrote:
> Do applications using rnp (e.g. Thunderbird) use this flag, though?

I wonder whether we're talking about RNP_LOAD_SAVE_PERMISSIVE or 
something else I'm not aware of.

Thunderbird (TB) uses flag RNP_LOAD_SAVE_PERMISSIVE only optionally.

In the early days of TB's OpenPGP support, the users wanted to migrate 
as many public keys as possible into TB's storage.

At that time RNP was very strict by default, and rejected a lot of keys. 
Therefore it was decided to use RNP_LOAD_SAVE_PERMISSIVE.

However, in recent versions, RNP_LOAD_SAVE_PERMISSIVE wasn't used by 
default when importing public keys. The initial import was always 
attempted without this flag. When the operation failed, TB showed a 
warning that something went wrong, and offered the user to try importing 
again in a more permissive mode (which turned that flag on).

I don't know how frequently users used that. And in the recent past, I 
concluded it's probably no longer necessary, as RNP seemed to be able to 
handle all commonly used keys.

We're currently working to make this flag more difficult to find. My 
thinking was, we probably shouldn't import keys that RNP doesn't know 
how to handle. [1]

I'm happy to test whether this flag works with the kind of keys we're 
discussing here, and see if Thunderbird can successfully use it with the 
known encryption key. (I'd appreciate a link to a public key that I 
could test, on- or off-list.)

If we find that it works, and if it were decided that v4-pqc keys shall 
be supported, we could consider to enable the flag, in the hope that it 
doesn't introduce other side effects. However, it would be preferable to 
get an updated version of RNP that fixes the intolerance problem.

Thanks
Kai

[1] https://bugzilla.mozilla.org/show_bug.cgi?id=1829881