Dear all,

Please find below the notes and action items from today’s meeting on Thursday, 8 December 2022 at 13:30 UTC.

Kind regards,

Ariel, Steve, and Emily

Notes and Action Items - IDNs EPDP Call – 8 December 2022

Action Items

Action Item 1: Staff to propose on list that beginning in January, calendar invites will be for 13:00 – 15:00 UTC with advance notice if the call needs to be the full 120 minutes.

Action Item 2: EPDP Team members to share on list if they have expertise or input that could assist with developing the risk assessment on options for string similarity.

Notes

Roll Call and SOI Updates

Welcome and Chair Updates

• Now that the Council has approved the proposed updated timeline, it’s important to meet this timeline. We’ll be going through the critical path today for the Phase 1 Initial Report.
• Leadership is asking for a commitment from the group in this final push towards the Phase 1 Initial Report.
• String similarity is an important item we need to get through to make progress on several other outstanding issues.

Review of critical path to Phase 1 Initial Report

• Slide 4 – Critical Path to Phase 1 Initial Report – the goal is to publish the Phase 1A Initial Report by April 21 2023.
  • The red box indicates charter questions that need to be closed. Some of these depend on the outcome of discussions on string similarity, so coming to a conclusion on string similarity is an important next step. The final items to cover are the “catch all” questions. Once language is drafted, it will need to be reviewed (pink box).
  • The blue box indicates items that have gone through initial deliberations. Staff and leadership will finalize draft text and send to EPDP Team for review and input.
  • The orange box indicates that the EPDP Team needs to review org input on specific charter questions.
  • In March, preliminary recommendations will be consolidated, and Initial Report will be built (orange and yellow boxes). The EPDP Team will have two sessions in Cancun during which the leadership team expects to make progress on bringing together the Initial Report.
  • Draft Initial Report will be shared on the mailing list in April, providing the EPDP Team an opportunity to identify any final missing items or other issues (purple box).
  • To get through some of this work in the coming months, calls may need to be longer or more frequent. As a reminder, we will be having a call on 22 December.
  • The expectation is that members will be consulting on an ongoing basis with their SO/AC/SG/Cs on the work of this group.
  • Question: Is there a reason for the SO/AC/SG/Cs to review the report prior to the public comment period?
  • Response: This is a question for the group’s themselves based on how they organize their work. When the draft report is published, groups can submit public comments.
  • When the leadership team sends draft text to the EPDP Team for review, it will remain conscious of providing enough time, where possible, for groups to consult with their groups on the language. The critical path anticipates four weeks of review for draft language.
  • There will be a two-week period to review of the whole report. Groups can incorporate feedback from their groups during that period, as well.
  • To the extent that groups have concerns about the timeline, the leadership team will seek to be responsive to the extent feasible.
  • The leadership team is available to provide a status update to the GAC if that is helpful.

ICANN Org input on the String Similarity Review hybrid model

• Slide 6 – Summary – Focus of Analysis, Overview, and Method
• Question: Isn’t Level 2 limited to Primary + Requested Allocatable, as opposed to Primary and All Allocatable?
• Response: It was Level 1 that includes Primary + Requested Allocatable, but on the slide, for simplicity, the assumption is that only the primary is being applied for.
• The number of TLDs selected are approximately in proportion to the those in 2012 round with respect to scripts.
• Slide 7 – Calculate the Number of Comparisons
• Number of Comparisons:
  • Level 1: 190
  • Level 2: 343
  • Hybrid: 13,003
  • Level 3: 95,144
• Slide 8 – Results, including Caveats and Takeaways
• Takeaways:
  • The mitigation of confusion risks will likely be enhanced from Level 1 to Level 3, but the cost of operating the String Similarity Review will likely increase from Level 1 to Level 3 
  • The added costs will likely be passed onto applicants, given the cost recovery principle 
  • Hybrid model is a compromise between Level 2 and Level 3 in reducing computational complexity involving blocked variants
• Comment: There are different ways to look at the numbers and the data is indicative only. Based on the numbers, the hybrid model is favorable compared to Level 3 but Level 2 is favorable compared to the hybrid model.
• From one perspective, it would be helpful to have examples of the specific comparisons taking place for the specific strings in the summary sheet.
• Question: For some strings the difference between the levels is quite significant, but the numbers are smaller for other scripts. Why?
• Response: Some scripts and some strings within those scripts will have more blocked variants. If these blocked variants are taken into account in the comparisons, there will be a spike in the overall number of comparisons.
• For this analysis, the staff team used a mix of scripts that is in proportion to the 2012 round. The labels themselves were chosen randomly to avoid introducing bias.
• Slide 9 – String Similarity Process & Hybrid Mode
  • Demonstrates where the hybrid model comes into play to help in understanding the operational and cost implications
  • This flow chart is built on the 2012 process.
  • As a reminder, the string similarity review is done by a panel, so more string comparisons will take more resources.
  • Note that the hybrid model may come into play in step 6 (string similarity review) and also step 8 (to the extent that it is used for the string confusion objection).
• Excerpt from org comment: “With added complexity, there is a high probability that the cost will also increase, and that those costs will be passed on to applicants given the cost recovery nature of the program.”
• Org also notes that increasing complexity of the review is associated with increased costs but also longer periods of time to complete the work.
• Given the inputs on today’s call are we comfortable moving forward with the hybrid model?
• Potential area for risk assessment – are the risks of failure modes likely/severe enough to justify the additional complexity and costs for applicants?
• Comment: It is important to be conscious of this balance noting the risks of increased application costs deterring applicants for IDN labels. A goal of our work is to promote IDNs.
• Additional consideration – In the future, the string similarity panel will likely need additional expertise to take into account that allocatable and blocked variants will be part of the review.
• One potential approach to the Initial Report is to include Level 2 and hybrid model as alternatives for public comment input.

AOB

• Reminder to use the email list to continue the deliberations between EPDP Team calls.
• Reminder that Donna will be based in Australia beginning in January. Suggestion to start calls 30 mins earlier and potentially keep the end time the same beginning in January. The calendar invites would be for 13:00 – 15:00 UTC, although that additional time would not always be needed.

Action Item 1: Staff to propose on list that beginning in January, calendar invites will be for 13:00 – 15:00 UTC with advance notice if the call needs to be the full 120 minutes.

Action Item 2: EPDP Team members to share on list if they have expertise or input that could assist with developing the risk assessment on options for string similarity.