I'm cleaning up and looking into committing the multiple crypt hook patch, but need some feedback from people who use it.
The current behavior (without that patch) is to prompt whether to use the crypt-hook value if one is found. If the user answers "no" then Mutt will use the original address for key selection instead. Since there can be at most one crypt-hook, this makes sense. With this patch, it's less clear what the correct behavior should be if the user user says "no". The current patch will use the original address and perform key selection with it, but doing this multiple times doesn't seem right. Would it be better for "no" to completely skip key selection *unless* it's the last crypt-hook and they haven't selected a key yet? As an example, let's say there are 5 crypt-hooks for an address. If you chose "yes" for the first crypt-hook, then you could say "no" to the next four and it wouldn't perform any key selection for those four. If you said "no" all five times, then for the first four, it wouldn't perform key selection, but the fifth "no" would cause Mutt to perform key selection on the original address. Does that sound reasonable, or would a different behavior be more useful? -Kevin
signature.asc
Description: PGP signature
