I'd like to request a reconsideration or at least clarification of this proposal.

According to the proposal, 35 specific candidate strings have been reviewed and only two strings `INTERNAL` and `PRIVATE` were found to satisfy the four selection criteria described in SAC113. However the proposal doesn't actually describe remaining 33 candidates, which greatly exceed the number of de facto private-use TLDs listed in SAC113.

The proposal also acknowledges ICANN Language Services in the assessments of the candidate strings in non-English languages. This might be a critical misunderstanding of SAC113, which clarifies that:

[...] Therefore, further strings could be considered to fit such criteria if and when such a need arises. For example, consideration could be given to internationalized versions of the string, reserving strings that are meaningful in different scripts.

I believe this is meant to translate the proposed string into different scripts, and not to possibly propose different strings for different scripts. Only two kind of assessments are possible in my reading of SAC113:

1. All candidate strings are in English, but they are additionally evaluated and filtered in order to not constrain the future need. Such evaluation is however provisional and doesn't block the final proposal.

2. Each candidate is specified as a set of strings in multiple languages, and the chosen set of strings is further distilled into a single proposed string. The cross-language evaluation is integral in this kind of assessments.

The proposal notes that `PRIVATE` in particular may carry unintended meanings in other languages, but otherwise the proposal is unclear about which assessment was used. This issue further strengthens a need for publishing all candidate strings in public.

Moving on, I agree that `INTERNAL` is a better label than `PRIVATE` for the reason described in the proposal, but I don't think it fully satifies the four selection criteria, particularly on the meaningfulness.

SAC113 doesn't really defined the meaningfulness, so IANA had to come up with their own definition which I found acceptable. The weakness of strings `DOMAIN` and `PRIVATE` was also (partly) attributed to this definition of meaningfulness, which helped my understanding of the proposal.

However it falls short of attributing the strength of `INTERNAL` in terms of meaningfulness. The intended meaning can be described as (A) "specific to an unspecified organization", thus implying it's (B) "unavailable beyond the organization's boundary". However the primary sense of the English word "internal" is only marginally related to the intended meaning, which is derived from the primary sense. If we only take the primary sense, the string `INTERNAL` can't necessarily imply (B) which seems necessary for meaningfulness.

In the list of popular private-use TLDs in SAC113, `HOME` and `CORP` were only other alternatives which uses are organically increased in some sense, while others can be traced back to specific pieces of software. Note that they directly point to the presumed usage instead of an abstract and vague "internalness". While they are clearly lacking in other criteria, their wild popularity should have been better noted at the very least.

It might be still true that `INTERNAL` is the best candidate string in spite of these issues. However my belief is that the current proposal doesn't adaquately justify that conclusion and therefore should be clarified and possibly reconsidered.

According to the proposal, 35 specific candidate strings have been reviewed and only two strings `INTERNAL` and `PRIVATE` were found to satisfy the four selection criteria described in SAC113. However the proposal doesn't actually describe remaining 33 candidates, which greatly exceed the number of de facto private-use TLDs listed in SAC113. They should be published in public for the better understanding and verification of the process.

I agree that `INTERNAL` is a better label than `PRIVATE` for the reason described in the proposal, but I don't think it fully satifies the four selection criteria, particularly on the meaningfulness. IANA had to come up with their own definition which I found acceptable, and the weakness of strings `DOMAIN` and `PRIVATE` was also (partly) attributed to this definition, however the same definition was never used to justify the strength of `INTERNAL`.

In the list of popular private-use TLDs in SAC113, `HOME` and `CORP` were only other alternatives which uses are organically increased in some sense, while others can be traced back to specific pieces of software. Note that they directly point to the presumed usage instead of an abstract and vague "internalness". While they are clearly lacking in other criteria, their wild popularity should have been better noted at the very least.

For these reasons, I believe that the current proposal doesn't adaquately justify its conclusion and therefore should be clarified and possibly reconsidered.