> -----Original Message----- > From: regext [mailto:regext-boun...@ietf.org] On Behalf Of Patrick Mevzek > Sent: Wednesday, April 18, 2018 2:45 AM > To: regext@ietf.org > Subject: [EXTERNAL] Re: [regext] WGLC: draft-ietf-regext-rdap-object-tag > > Pieter, > > On Mon, Apr 16, 2018, at 15:39, Pieter Vandepitte wrote: > > No, indeed it isn't about discovering entities, but discovering the > > authoritative server for an entity, basically wat is meant with > > bootstrapping. > > [..] > > > So my final vote on this draft: if there's a need for "globally unique > > handles" (I'm not convinced they are needed to help the clients), then > > yes, go ahead, but I would prefer a hyphen (it works, as long as the > > repos do not have hyphens), the same as roid in EPP. > > > FWIW, note that I completely share both your observations and your > conclusions. > So that you do not feel alone :-), we are at least too, but clearly the > minority. Such is life. > > > More broadly, I am doubly sad that hyphen is rejected because already seen > elsewhere. > > First, because RDAP for domain names has not reached critical mass and far > from it in fact (newer discussions on GDPR should have ressuscited it > since it is clearly suited for it, but things being done at the last time > did not make it possible I guess to think about it, and I applaud Scott's > energy and experiments on authentication models for it), so no characters > should be ignored just because some other registries did start to use it. > When IDNs started, before IDNA became a standard, names were sold and used > with the bq-- prefix. This was absolutely not a reason not to choose a > "proper" prefix, even if it meant at the time to break any existing bq-- > domain names
Remember, though, that the situation with bq-- is precisely why a decision was made to use xn-- instead. > Secondly, not using hyphens will make it very hard for anyone to convince > me that there is really an overlap between EPP and RDAP (which was a core > justification to create this WG), as hyphen is clearly the choice in EPP. > > The rdapConformance part could have a "objectTag-0" token whose presence > would signal the client that the hyphen in object IDs has a special > meaning because it is a separator. This would allow current registries > using hyphens not as separator to continue working in the same way as they > would not present this "extension" in their response. Client would just > need to adapt, but it is the case what ever happens and characters is used > since there is no guarantee that all IDs, even in the same repository, > will be formatted with it. This is worth exploring. What do people think about going back to HYPHEN MINUS as a separator and noting the significance of the character with an "objectTag-0" token in the rdapConformance array? The draft should probably include text requiring addition of an "objectTag-0" token in the rdapConformance array to note support for this practice no matter which separator character is used. Scott _______________________________________________ regext mailing list regext@ietf.org https://www.ietf.org/mailman/listinfo/regext
