Long live... eh, I hope the new KSK won't live as long as the old one.
Anyway, it doesn't really matter how fast people update this.
https://www.ietf.org/mail-archive/web/dnsop/current/msg24989.html
See RFC 5011 for details of the protocol.
I re-tested validation with both of these files, to be sure.
Nitpicks:
- The timestamps there were useless.
- The generator now switched the two keys; I don't know why.
I intentionally remove the comments like "state=1 [ ADDPEND ]".
The problem is that keys e.g. in ADDPEND state are *not* immediately
usable for validation - see RFC5011 for details. I verified that Unbound
does disregard this on the format we and Debian use ATM, presumably due
to removing parts of the comments, but it would be confusing nevertheless.