> -----Original Message----- > From: regext [mailto:regext-boun...@ietf.org] On Behalf Of Thomas Corte > Sent: Monday, February 13, 2017 9:28 AM > To: regext@ietf.org > Subject: [EXTERNAL] Re: [regext] Query: Remove contact:postalInfo from a > contact > > Hello, > > On 13/02/2017 14:59, Hollenbeck, Scott wrote: > > > Jim, the change option should be interpreted as "completely replace > > old value with this new value". Whatever postalInfo is provided with > > an update should completely replace whatever was there before. Looking > > at the use case you described, yes, you can replace > > <contact:postalInfo type="int"> with <contact:postalInfo type="loc"> by > including the "loc" > > value in the <contact:chg> element. If that's not clear it's only > > because I didn't think the meaning of "change" was ambiguous. That's > > why there's nothing there to selectively identify the values to be > altered - "change" > > means "completely replace". > > If the idea were to completely replace all postal info by whatever > <postalInfo> elements are specified in the <chg> element, then there would > be no point in having a special - potentially completely empty - > "chgPostalInfoType" type in the schema there, since it allows the addition > of incomplete address data that could result in schema violations for > <contact:info> responses. This could just be "postalInfoType" then to > enforce the presence of mandatory fields. > This seems to indicate that the original idea was *not* a complete > replacement. This would otherwise also kind of contradict the way <chg> is > interpreted for domains (RFC 5731), since it would imply that one could > not change a domain's registrant without also changing or removing a > domain's authinfo string.
My apologies. I wrote the text more than 15 years ago, and after looking through the documents again I see that you're correct. As written, the schema and text support selective change. I'm not sure why or how we ended up there, but that's what we have. At least I'm consistent in my memory of my intentions: https://www.ietf.org/mail-archive/web/provreg/current/msg07561.html So back to the original question: I believe it's possible to remove one of the forms by replacing it with an empty value as described here: https://www.ietf.org/mail-archive/web/provreg/current/msg07568.html > <contact:update xmlns:contact="urn:ietf:params:xml:ns:contact-1.0"> > <contact:id>C20131114-01</contact:id> > <contact:chg> > <contact:postalInfo type="int"/> > </contact:chg> > </contact:update> It looks schema-legitimate to me since all of the child elements of <contact:postalInfo> are optional. Looking at all of this again I wish I had just used the same schema type for creates and updates and explicitly noted that a change is a 1-for-1 replacement. That would seem to be much simpler. Scott Scott _______________________________________________ regext mailing list regext@ietf.org https://www.ietf.org/mailman/listinfo/regext
