US National Body

Comments on JTC 1/SC2 N3438

Application for Registration No. 230, Turkmen Alphabet


We see the following problem with the proposed registration.


The Description field states "A set of 94 graphic characters constituting a version of ISO 8859".  This entry MUST be corrected.


ISO/IEC 8859 does not have an concept of "versions" -- it has different parts, such as 8859-1, 8850-2, etc. each one of which is a standalone standard.  There is no 8859-xx containing the proposed G set from Turkmenistan.


If it is meant to be "similar" to other 8859 parts -- per provisions in ISO/IEC 4873 -- i.e. an 8-bit code with left half (G0 set) as ISO/IEC 646-IRV, and the proposed set as the G1 set, then the proposed G set has DUPLICATES with ISO/IEC 646-IRV and is unsuitable for that context.  Also, a G0 ESC 02/08/02/01 F -- cannot be assigned in this context.


If the registration is meant to be a version of ISO/IEC  646, then this is again unsuitable because it changes the INVARIANT positions of ISO/IEC 646.


The only possible way this set can be registered is as an independent G set under the provisions of ISO/IEC 2022, and the corresponding registration procedure ISO/IEC 2375.  The Description field has to reflect this.


We also suggest that the names of characters be the same as (aligned with) those in ISO/IEC 10646, and a column be added to show the corresponding UCS code positions -- per proposed revision to ISO/IEC 2375 that is under consideration in SC2.