Dear Kevin, Am 04.07.2026 schrieb Kevin J. McCarthy: > > if (key->expired) > > flags |= KEYFLAG_EXPIRED; > > > > to the corresponding S/MIME loop and it works as expected (although the > > option pgp_show_unusable seems to be ignored but I have to further check > > that). > > I'm glad that is working. That's good in that it means the S/MIME GPGME > backend flags are set. So it's just a matter of hooking them up.
OK. So I can produce a patch for that, however:
> The check for the $pgp_show_unusable is there in crypt_select_key(), so
> I'm not seeing right away why that isn't working for you. You may have
> to add muttdbg() statements and see what the problem is.
I tested all variants that came to my mind and I get this result:
- with check for KEYFLAG_EXPIRED in S/MIME code:
- $pgp_show_unusable = no
- S/MIME: no choice
- GPG: no choice
- $pgp_show_unusable = yes
- S/MIME: no choice
- GPG: no choice
- without check for KEYFLAG_EXPIRED in S/MIME code:
- $pgp_show_unusable = no
- S/MIME: menu shown
- GPG: no choice
- $pgp_show_unusable = yes
- S/MIME: menu shown
- GPG: no choice
So it seems that $pgp_show_unusable has neither effect for GPG nor
S/MIME but the check for KEYFLAG_EXPIRED has an effect.
I did not see any messages regarding that in the debug log so I have to
check the source code (or run an actual debugger). For now, that has to
wait a while.
> > %[<s>] date of the key where <s> is an strftime(3) expression
> Yeah, that's not going to be changeable. The format string is shared across
> GPGME and PGP classic, and I assure people will notice if you just change
> it. Adding multiple dates also involves mucking in the gnupgparse parser,
> which I'd rather not do for something like this.
Thank you very much for approaching this extension of $pgp_entry_format.
It looks as if it was not too difficult to implement. I'll test it.
Best,
Robert
signature.asc
Description: PGP signature
