On Sep 23, 2015, at 4:40 PM, Dave Lawrence <t...@dd.org> wrote: > Ted Lemon writes: >> It would be helpful if the authors could explain why the REFUSED >> response is being used here. > > Not to be glib, but because that's what Wilmer originally specified. > That's thus what got implemented by the existing implementations (and > there are more than you'd likely imagine, too).
Thanks, that’s what I expected. My concern is that the document currently states this unapologetically, so a reader who does not know the context might be tempted to implement it that way. The way the document is written at present is as if it is a protocol specification intended to be implemented. There is no applicability statement that says something like "there are a lot of problems with this version of the spec, and you should wait and implement version 2 if you don’t already have an implementation or need to interoperate with version 1." I think the document needs to have some clear language restricting its applicability, and I also think that in cases where what was done in version 1 was obviously wrong, the document should say so. I think it’s actually harmful to publish the version 1 document as an RFC without the version 2 document, to be honest. I get why you want to do it in two steps, but from where I’m sitting this feels like an unfortunate decision if the first step is to publish something that looks like a perfectly valid spec and then the spec that does things the right way comes possibly much later on.
_______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
