Current admin division of the country
In my experience the admin division in Iraq looks something like this: 
What is currently marked as 'level 9 - administrative districts in Baghdad' is actually an old qadha/district (admin_level = 6) division. What is currently marked as 'level 10 - neighbourhoods in Baghdad' is consistent with Wikipedia, but unknown to any Iraqi official I've spoken to. There may be areas established ad-hoc around 2003 that are not really used by the government.
I think it should be worth adjusting the OSM model Tag:boundary=administrative, while preserving the current tiers as much as possible.
Hello, I copy and paste the response that I made to mail of Øukasz:
In Iraqi Kurdistan they speak several Kurdish dialect, Kurmandji (kmr), ckb (Sorani) and hac (Gorani).
Kurmandji (kmr) is written in the Latin alphabet and not in Arabic script so the kmr names must be written in Latin. We can do a transliterated using the code name:kmr-Arab but not necessary.
Sorani (ckb) is written in Arabic scipt; so names should be written in Arabic script. We can also using name: ckb-Latn for transliteration.
- Geographical distribution of name:ku
- Geographical distribution of name:ckb:
- Geographical distribution of name:kmr:
Actually name: ku has been used for the dialect Kurmanji (name:kmr) for the names of cities and countries in the world. For Sorani is the ckb.
It is generally used for the Latin alphabet. Therefore, I propose to always use it with the two other (kmr and ckb) but only in Latin:
- For cities or population mainly speak Sorani (ckb); use name: ku for transliteration Sorani in Latin.
- For cities or population mostly speaks Kurmandji (kmr); use name: ku for the Kurmandji.
This allows us to have a map of Kurdish regions with a single ku code readable by all Kurds. To see the details of the dialects we can always refer to name: kmr and name: ckb.
- In others countries:
In Syria they speak only Kurmandji : name: ku fact refers to Kurmandji.
Turkey talking Kurmandji (kmr) and Zazaki (diq) and name: ku mainly refers to kmr (there is no ambiguity). Sorani is not spoken.
Iran is spoken Kurmandji (kmr), Sorani (ckb), Southern Kurdish (sdh), hac (Gorani).
Hello I just look for the transliteration ckb-Latn:
When writing the Sorani into Latin alphabet; you must use the Kurdish alphabet.
This link allows you to switch from Arabic alphabet to Latin Kurdish alphabet.
For example (correction):
- name:ckb = قرةقوش
- name:ckb-Latn = Qirequş
- alt_name:ckb-Latn = Qerequş (?); Qereqoş (?)
- name:kmr = Qerequş
- alt_name:kmr = Qereqoş
- name:ku = Qerequş
- alt_name:ku = Qereqoş
- The 'ku' code refers to a macrolanguage, so if it were used for one of the Kurdish family languages ambiguity would inevitably creep in. Why not use kmr for Kurmanji for simplicity?
- Also, regarding the Kurdish alphabet - it would be great to have standardized spelling, however I do not believe the transliteration system you are linking is really often used in Iraqi Kurdistan. For example, I have only come across the spelling you propose - 'Qerequş' - in Wikipedia, and spellings not using Kurdish diacritics are more common (including Rudaw that you are citing as well as national offices). --Øukasz (talk) 21:02, 7 July 2017 (UTC)