Public Comment is a vital part of our multistakeholder model. It provides a mechanism for stakeholders to have their opinions and recommendations formally and publicly documented. It is an opportunity for the ICANN community to effect change and improve policies and operations.
If no, please explain why.
I see a serious problem with this document: the work done on the Root Zone is being transferred wholesale to the Second Level Domain Names. Why is this a problem? Because the two situations are not identical. In particular, proposed new gTLDs are subject to a manual review, to assure that they will not cause confusion. In contrast, Second Level Domain Names are not (and, realistically, cannot be) subject to any kind of manual review. The script generation panels efforts reflected this. In the Latin GP, we were told explicitly and repeatedly that we should NOT consider the Second Level when doing our work. In recognition of this, when the Latin GP was identifying variants for the RZ-LRG, we set a high threshold for identifying variants – exactly because new gTLDs would get a manual review. When we were contacted last year by the authors for comments and advice on the Second Level, the Panel recommended that the threshold for variants should be lowered for the Second Level. Specifically, we suggested that the threshold be lowered from 5 of the 7 members finding a pair confusable, which was used for the RZ-LRG, to having 3 of the 7 members doing so be sufficient for the Second Level. (This should not require a serious effort on the authors’ part. The spreadsheets where the individual members of the Latin GP evaluated pairs of symbols still exist. All that is necessary is a very low level clerical effort to review those spreadsheets. No expertise or judgement required.) Unfortunately, the authors appear to have decided not to follow this recommendation. Perhaps an example of the problem would be helpful. Of the 200+ languages reviewed by the Latin GP, a handful include Unicode member 00ED, Latin Small Letter I with Acute (í). The Latin GP decided that this would not be a variant of Latin Small Letter I. (Note that this was a very close decision. Witness the fact that letters with a dot above, and letters with an acute, are not variants if the underlying letter is a vowel, but are variants if the underlying letter is a consonant.) So consider. Suppose someone submits a proposal for a new gTLD of .míl. It seems fairly certain that the manual review would reject this out of hand as confusingly similar to the longstanding .mil TLD – users who have not encountered one of the languages which use the I with Acute (i.e. the vast majority of users) would simply assume they were seeing a font feature, assuming that they noticed at all. On the other hand, if someone wants to register ícann.org or cítí.com, under the rules envisioned by the document, both of those would sail right through. Now ICANN might not like having someone else control the variation on its domain name, but the amount of potential damage from it is limited. On the other hand, someone other than Citi Bank controlling the variation of their domain name could be an enormous problem. I suspect that Citi would not be amused. The potential for DNS Abuse, not just confusion, is obviously huge. In short, the approach envisioned by the document seems certain to result in unnecessary confusion for Internet users. And, in many cases, provides larger scope for DNS Abuse than necessary. I strongly urge the authors to go back and reconsider their approach to variants when it comes to Second Level Domain Names.
For the Latin Script LRG, the set of variants used for the Second Level needs to be seriously expanded. Specifically, the document should follow the recommendation of the Latin GP to lower the threshold for variants from that used for the Root Zone.
It is my view that it is inappropriate to use the set of variants used for the Root Zone in the RZ-LRGs for the Second Level.
The Root Zone LGRs were developed under the explicit mandate to ignore any possible implications for the Second Level. This matters because proposed gTLD registrations are subject to manual review, to avoid confusion. Whereas Second Level Domain Name registrations are not, and cannot be, manually reviewed. The Latin Generation Panel accordingly recommended, when the authors asked last year, that a lower threshold for variants for Second Level names should be applied. This does not require a large effort, merely a clerical review of the existing evaluations from the Latin GP's Root Zone efforts.
In the interest of reducing confusion, and of avoiding enabling DNS Abuse, I strongly recommend that the authors revise the document to reflect that suggestion.