The call for the IDNs EPDP team will take place on Thursday, 17 March 2022 at 13:30 UTC for 90 minutes.

For other places see:


  1. Roll Call & SOI Updates (2 minutes)
  2. Welcome & Chair Updates (5 minutes)
  3. Recap on Charter Questions  D1b (including proposed survey questions for eligible ROs), B4a, and B5 (80 minutes)

            Working documents

       4. AOB (3 minutes)


EPDP Team Meeting #26 Slides - D1b, B4a, B5.pdf


Audio Recording

Zoom Recording

Chat Transcript 

GNSO transcripts are located on the GNSO Calendar



Apologies:  Joseph Yee

Notes/ Action Items

Action Items:


Action Item #1: EPDP Team members to provide input on the best way to conduct a survey with ROs located in China.

Action Item #2: Staff to develop draft text for Part 1 and continue to develop questionnaire for Part 2. For Part 3, there appears to be convergence on the cost recovery principle which could be a basis for draft text.

Notes – IDNs EPDP Call – 17 March 2022

Welcome & Chair Updates 

  • The focus of today’s call is a recap of charter question D1b, as well as deliberations on B4a and B5.
  • Thanks to those who provided input on draft responses to charter questions A4, A5, and A6. These comments will likely be discussed during next week’s call.

Recap on Charter Questions  D1b (including proposed survey questions for eligible ROs)

  • The purpose of the recap is to confirm that there is agreement on the general approach to this charter question, so that the leadership team can take the input back to draft language for EPDP Team review.
  • Slide 4 – summary of discussion for Part 1: A new applicant seeks to apply for a new gTLD and allocatable variant label(s) of that gTLD.
  • Comment: We see efficiencies when we manage to process all of a variant set “in one go.” If a label in a set happens to be in contention with another label that has been applied for, but the applicant is not deciding to activate that variant at that time, how will this be handled? There is a charter question that specifically addresses strings that are held for an RO, so this scenario may be addressed when that charter question is discussed.
  • Would it be helpful to visually depict or map these types of use cases?
  • Additional consideration: there could be scenarios where an applicant is applying for a single string and variant labels are generated through the RZ-LGR that may not be desired by the applicant. If all allocatable variants are being processed, there may be additional fees “imposed” on the applicant even if they do not want to use these variants.
  • Clarification question: it seems to be the case that if an applicant does not want the variant, it does not need to include them in the application and therefore they will not be evaluated.
  • Two scenarios: 1. We are pre-evaluating all variants, which means we are treating all variants as if they could be used. 2. We only pre-evaluate variants that the applicant is asking to be evaluated.
  • Recollection that the team seemed to previously agree on pre-evaluating only those variants that the applicant has requested to be evaluated.
  • Comment: there may be legal challenges if one entity wants variants in the future that another entity is entitled to activate under the same entity principle.
  • Comment: circumstances change over time. This should be taken into account in considering the possibility of activating variants after a period of time has passed since the original string is applied for.
  • Additional consideration: looking at all variants, not just those applied for, might require evaluating very large numbers of strings in Arabic.
  • Question: is it correct to assume that if you don’t apply for variants all at once, you can request to activate additional variants in the future?
  • Response: it seems logical for an applicant to ask for some variants initially and then ask for additional variants later based on their experience.
  • Staff input: as far as the number of allocatable variants, staff ran a list of TLDs that are already delegated. For Han script, it could go up to 18 variants for the delegated TLDs. For Arabic delegated TLDs, the largest number of variants is 12,400.
  • Consideration: it is helpful to consider the likelihood of overlapping variants between two applied for strings that are independent. If there are “hidden” variants that are not considered in the process, the first come first serve rule is not fair in the objection process.
  • Response: if there are two labels with variants that overlap, then the two primary labels would also be variants of one another as well.
  • Question: when we think about an IDN gTLD and its variants, are we considering them as a set or as separate strings? From what we are hearing, we seem to be discussing them as separate strings. If considered as a set, is it easier to deal with?
  • Comment: it can be considered a set in cases where an applicant applies for a primary string and its variants all at once. It can also be considered separate if a TLD is applied for in one round and then the RO wants to activate/apply for a variant later.
  • Comment: an applicant should show it can handle the variants it applied for. It should not need to show it can handle all possible variants.
  • Comment: across rounds, the first to file is the one who gets the string along with all variants where they express interest in those variants or not.
  • Slide 5 – summary of discussion for Part 2: An existing RO seeks to activate allocatable variant label(s) of its existing gTLD.

Action Item: EPDP Team members to provide input on the best way to conduct a survey with ROs located in China.

  • Comment: it is important not to take results of one meeting and consider that an agreement. Proposal to revisit the idea of conducting a survey.
  • Comment: the EPDP Team should not ask ROs about fees in a survey. ROs will always want to pay lower fees. SubPro has already addressed how fees should be determined for future applications. It did not differentiate variants. We should affirm the principle of cost recovery. We shouldn’t propose any fees unless we understand what the evaluation will cost. Important that we don’t limit the scope of the issue to existing ROs from the 2012 round who want to apply for a variant in the future.
  • Clarification: purpose of the questionnaire is to get a sense of the interest of existing IDN gTLD applicants in applying for a variant if that option becomes available. The survey will be limited and focused with the intention of helping the EPDP team address this charter question.
  • Regarding whether this question is limited to ROs from the 2012 round or whether it also includes ROs who apply for primary strings in the future and then later seek variants, staff recalls that this question was limited to 2012 ROs. It was confirmed by Chair of charter drafting team that there are two different questions to address.
  • Comment: these questions are connected to the question of same registry agreement or different registry agreements and should be considered in a matrix format with respect to fees.
  • Comment: we can handle TLDs differently when they applied for the original string in 2012 compared with those who apply in the future where there will be the possibility of also applying for variants.
  • From another perspective, there should be one process that applies to ROs from 2012 and future ROs who later want to activate variants, so that you don’t give unfair advantage to existing ROs.
  • Clarification: The 2012 round has unique characteristics because the round has already passed and applicants did not have the option of applying for variants. Certain components should be repeatable in the future for future processes. We need to look through two different lenses although a common solution may end up being applicable. 
  • Comment: we may not be able to blend the two. Support for collecting additional information in a survey to support decision-making on applicants from the 2012 round.
  • Comment: the EPDP Team needs to finalize requirements before we can talk about fees. We should agree to recommend a process by which a 2012 RO can activate variants during a limited period of time.
  • Returning to the purpose of the survey: the goal is to get a sense of the demand among existing ROs from the 2012 round to activate variants.
  • Suggestion: use the term “activation” for variants obtained outside of a round as opposed to “application.”
  • Slide 6 – summary of discussion for Part 3: Associated fee(s) and specific implementation guidance.
  • Comment: we should not get into formulas or numbers due to anti-trust concerns. We should focus only on principles. SubPro came up with a formula that is essentially cost recovery. This principle should apply to our work, too.
  • Clarification: this team is trying to establish what we think is a reasonable path forward. The principle of being revenue neutral is understood by this group. For future rounds, we have agreed that an applicant can apply for a TLD and its variants together. There may be additional process to evaluate this application. Should there potentially be additional loading to accommodate that?
  • Comment: if the EPDP Team agrees that if an applicant wants a variant and there are increased costs for that evaluation, there should be additional fees to cover that evaluation.
  • Response: it may not be as simple as it looks. There are many different possibilities about the number of variants included in an application. Would there be an additional fee per variant?
  • Reminder: the purpose of applying for the variant is to ensure that the end user has the same experience even if they type the variant. It is the experience of a single TLD.
  • Comment: if it costs more to evaluate more variants, the application should cost more. ICANN should be determining the exact calculations, but cost recovery should apply.

Action Item: Staff to develop draft text for Part 1 and continue to develop questionnaire for Part 2. For Part 3, there appears to be convergence on the cost recovery principle which could be a basis for draft text.


  • Slide 8 – Context for charter question B4a
  • Suggestion: this is a good place to say that just because a string is “withheld same entity” the RO does not have any rights to those names outside of the approved request process.
  • Response: In a situation where strings went through an evaluation process and then have the designation “withheld same entity” there might be legal claims to that string.
  • Question: Would “withheld same entity” status only apply to those that have already been included in an application and passed evaluation or would it also include those that are not included in the application and withheld?
  • Response: We have not yet determined the answer to this question.

  • No labels