Summary: | kontact/kmail does not encrypt correctly with gpg when a BCC: is set | ||
---|---|---|---|
Product: | [Unmaintained] kmail | Reporter: | David Guembel <david.guembel> |
Component: | encryption | Assignee: | kdepim bugs <kdepim-bugs> |
Status: | RESOLVED DUPLICATE | ||
Severity: | major | ||
Priority: | NOR | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Gentoo Packages | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
David Guembel
2004-10-19 18:06:31 UTC
> From: David, To: Alice, BCC: David
> ..should be encrypted with the keys of both Alice and David, but only gets
> encrypted with David's key.
That's correct. The point of BCC is that the recipients in To: and CC: do not know that the recipients in BCC are receiving the message. Having the email encrypted to the BCC recipients would defeat this. Kmail 1.6.1 solves this by sending seperately encrypted copies to the BCC recipients. (So in the above example, Alice would get the mail just encrypted to Alice and David would get the mail encrypted to Alice and David). Is it different in kontact/newer kmail?
Yes, I know about the problem of unwanted revealing of BCC recipients to others, and I have seen that KMail/kontact sends two messages. This worked perfectly until kde 3.3.0, but with 3.3.1, the message that arrives at the remote destination on the To: (i.e. Alice in the above example) is not encrypted with Alice's key at all, but only with David's. This can't be correct behavior ;) I think this is rather major. I can't reproduce the problem as it's described in the bug report. Therefore I think this bug is in fact a duplicate of bug 92412. If that's not the case then we need an exact description how to reproduce the problem (including a key for Alice which we can use for testing). Thanks for your comment and for your work! I shall look into this and keep you posted. In the meantime, would it be possible that you generate a patch against the kdepim-3.3.1 tarball, so I can test if that patch makes the problem definetly go away? OK, it seems there was a misunderstanding between me and the recipient of the messages of the type explained above I sent. So very probably, this is really a duplicate of bug 92412, as my recipient recieves empty messages for which he is required to enter his passphrase in order to view them. Belase note, however, that the messages (of the type specified in this report) I sent to him were also signed, but arrived encrypted (and empty) but unsigned with him. I would be happy to test the patch you submitted to CVS if you made a version of it against latest kdepim source tarball (3.3.1), and if that fixes the problem, I personally think this report can be closed. |