On 14.12.22 10:19, Joe Abley wrote: > Hi Martin, > > On Wed, Dec 14, 2022 at 10:08, Martin Schanzenbach <mschanzenb...@posteo.de> > wrote: > > > "Developers are wholly responsible for dealing with any collisions" > > > > I think this is an impossible task and as a developer that is addressed > > here I have to say that we cannot do that unilaterally for our > > implementation/design because collisions occur when _others_ do > > something. > > I don't understand why you say this is impossible when it is what happens > today and what has always happened. > > But perhaps I don't understand something about how "developers" is intended > to be interpreted. I agree it would be nice to clarify this sentence if it is > to remain. >
I think my main issue is the word "wholly". The developer cannot be "wholly" responsible. I can choose a label (e.g. "foo.alt") that is not already taken right now. But I cannot really do anything if somebody else comes along and uses the same label. It is not my responsibility to mitigate the emerging collisions. Or how is it expected that I do that? Change my design? Bribe the other group? "bob.foo.alt" still squarely falls into "my" namespace, so I cannot write a single line of code to mitigate the collision. Not only is it not really possible for me as a developer to do that, it is also unlikely that I even notice this issue as a user. Is the developer of a DNS server/resolver in any way responsible if "alice.eth" collides with ENS? Or what do you mean by "this happens today"? BR Martin > Joe > > > _______________________________________________ DNSOP mailing list DNSOP@ietf.org https://www.ietf.org/mailman/listinfo/dnsop
