Hi Dennis, On 25.07.2022 21:46, Tan Tanaka, Dennis via Gnso-epdp-idn-team wrote:
A few additional comments on 2.3 re: use of “primary label”
* Some questions I reckon we will face throughout: what is the “primary” label? what function does it serve? should we use such nomenclature? … some background notes: * When we talk about variants in a set, it is important to note which of all labels is the source for variant label calculation, or as we have adopted: the primary label. As a way of context, first, the variant relationship between two code points must satisfy the symmetrical property (there are other properties, but not relevant to this topic). This means, if ‘A’ is a variant of ‘B’, then ‘B’ is a variant of ‘A’. Second, any variant relationship is also assigned a disposition value: allocatable or blocked. The disposition value is unidirectional. Which means, a variant relationship could yield an ‘allocatable’ variant in one direction (e.g., AàB: allocatable), but a ‘blocked’ variant in the other direction (e.g., BàA: blocked). These rules are encoded in the LGR. * As we progress in our deliberations about the lifecycle of gTLDs it is important to understand (and remember) how the various variant labels and disposition values came to be (to discern which ones are allocatable/withheld or blocked). However, I believe we need to remain open (or at least give thoughtful consideration) to cases where one TLD label in the set needs to be handled separately e.g., retirement, and the implications of this one change applies to the rest of the labels in the set. For instance, what if the registry operator wants to retire the “primary” label and continue to operate the rest of variants. In this case, it would be unwise, perhaps, to re-assign the attribute of “primary” to another label in the set, because that might change the profile of the set as a whole (of course, this would depend what “primary” means and how it is used).
This is a very good point. Especially with asymmetric disposition we have a real problem if a primary label is retired and the other variants are kept. There might arise a situation in which none of the remaining variants can be made primary, because it would entail one of the other existing variants to become blocked instead of allocatable. The question would then be * Do we allow a variant set to exist without a primary label? * Will retiring a primary label require the registry to select another primary label and thereby possibly force them to retire other variants together with the former primary label? Take the following quite simple example. Applied for primary label: bıß Allocatable Variants: biß, bıss, biss So, someone might have all four labels in the Root zone at some point. If they now retire the primary label bıß, the other three labels could not exist at the time (without bıß also existing). Case 1: biß is made new primary Consequence: bıss becomes a blocked variant. Case 2: bıss is made new primary Consequence: biß becomes a blocked variant. Case 3: biss is made a new primary Consequence: both bıss and biß become blocked variants. Even though the example is not very likely, it's certainly possible and in other scripts might be similar examples that are more likely. So, to summarise my points: * Initially, we need a primary label to determine the disposition value of all variants. * Retiring a primary label may cause problems and we need to decide how to deal with these problems. Best regards, Michael -- ____________________________________________________________________ | | | knipp | Knipp Medien und Kommunikation GmbH ------- Technologiepark Martin-Schmeisser-Weg 9 44227 Dortmund Germany Dipl.-Informatiker Fon: +49 231 9703-0 Fax: +49 231 9703-200 Dr. Michael Bauland SIP: Michael.Bauland@knipp.de Software Development E-mail: Michael.Bauland@knipp.de Register Court: Amtsgericht Dortmund, HRB 13728 Chief Executive Officers: Dietmar Knipp, Elmar Knipp