It appears that Mark Andrews <ma...@isc.org> said: > >What is the reasoning behind the following? Why not just FORMERR the request >when the option length is not zero? How hard do we think writing a client is >that they will get sending a zero length option wrong? What is wrong with >sending back an immediate error signal? There is a thing of being too over >permissive which is why we have so many issues with the DNS today.
By golly, he's right. It would be better simply to remove any text about how to respond to invalid options. We write specs to say how to interoperate, not how to work around bugs we guess other people might have. R's, John _______________________________________________ DNSOP mailing list -- dnsop@ietf.org To unsubscribe send an email to dnsop-le...@ietf.org
