Petr Špaček <[email protected]> writes: > Does it make sense now?
Yes, that was a good set of helpful examples. Make sure that in the document (it may be there already -- I didn't check) you also cover all the cases: what should a resolver do when it hasn't been upgraded and it sees a new type within the range? Is it still safely able to proxy RRTYPEs not in the range (like it always has before with TYPEfoo)? I'd hope all range categories can adequately explain what each type of software must do when it sees a value in a range it doesn't understand. Excellent if it can "don't panic and keep [function]ing"! Thanks for the explanation. I still have reservations that needing 3 (or 4) ranges means we may not be predicting all future cases well enough, but that's up to future-us to fix. -- Wes Hardaker Google _______________________________________________ DNSOP mailing list -- [email protected] To unsubscribe send an email to [email protected]
