[Sks-devel] unwanted tolerance of buggy keys

[Clint Adams]

This key

http://zimmerman.mayfirst.org:11371/pks/lookup?op=get&search=0xED34CEABE27BAABC

is buggy.  It contains a generic certification packet on the first subkey
and a positive certification packet on the second subkey.

>From a quick glance at the SKS source code, it looks as though the signature
type is not being checked.

If I read RFC4480 section 11.1 correctly, the only signature types valid on
a subkey should be 0x19 and 0x28.

Could you please implement this restriction in SKS?