[Base] Incorrect Name Selection for Indian Ocean

[Ian Wagner (ianthetechie)]

Problem Description

Release: 2026-07-22.0
Theme: Base
Type: Water
GERS ID: b3e971b7-cd6d-3511-b5aa-e52afcec61f7
Original source (exact history version noted in the sources): https://www.openstreetmap.org/node/305640074/history/54

The Indian Ocean's name is reported as "Indischer Ozean" which I believe is confusing at best for the following reasons:

• The original source (OSM) has no name key
• OSM contains more than one name:* language-specific variant (79), so "primary" is ambiguous
• The selected fallback is not a widely used "international" name (German speakers aside :D)

I believe the Overture processing pipeline is selecting either name:als (lexicographically ordered first?) or name:de (intentional but odd fallback?). It is not currently possible for downstream consumers to tell what is going on without further documentation or access to the pipeline source code.

Possible Resolutions

In the case that there are multiple name:* variants without a name tag, I would suggest the following behavior:

• Acknowledge that there is no clear primary name and make the field nullable.
• Have an explicit (documented) fallback order, or else null.
• Have an explicit (documented) fallback order, or else choose the first via some deterministic ordering.

These are ranked roughly in order of my personal opinion of correctness, but I realize that my first 2 suggestions is a breaking schema change. As far as I can tell, if names is present (it is itself optional 🤔), then primary is a non-optional stripped string. In case Overture elects to use an explicit fallback order without making names.primary nullable, it would be nice for consumers to know when a fallback is selected / the result is ambiguous, so downstream consumers can make an informed cartographic decision.

Supplementary data

Here are JSON dumps of a few fields for quicker reference.

names

{"primary": "Indischer Ozean", "common": {"als": "Indischer Ozean", "ar": "المحيط الهندي", "arc": "ܐܘܩܝܢܘܣ ܗܢܕܘܝܐ", "azb": "هیند اوقیانوسو", "be": "Індыйскі акіян", "be-Latn-tarask": "Індыйскі акіян", "bg": "Индийски океан", "bn": "ভারত মহাসাগর", "br": "Meurvor Indez", "ca": "Oceà Índic", "cdo": "Éng-dô-iòng", "cs": "Indický oceán", "csb": "Jindijsczi Òcean", "cv": "Инди океанӗ", "da": "Indiske Ocean", "de": "Indischer Ozean", "dsb": "Indiski ocean", "el": "Ινδικός Ωκεανός", "en": "Indian Ocean", "eo": "Hinda Oceano", "es": "Océano Índico", "eu": "Indiako ozeanoa", "fa": "اقیانوس هند", "fi": "Intian valtameri", "fr": "Océan Indien", "gan": "印度洋", "gd": "An Cuan Innseanach", "gn": "Paraguasu Índiko", "grc": "Ἰνδικὸς Ὠκεανός", "hak": "Yin-thu-yòng", "he": "האוקיינוס ההודי", "hi": "हिन्द महासागर", "hr": "Indijski ocean", "hsb": "Indiski ocean", "hu": "Indiai-óceán", "hy": "Հնդկական օվկիանոս", "hyw": "Հնդկական Ովկիանոս", "ia": "Oceano Indian", "io": "Indiana oceano", "is": "Indlandshaf", "it": "Oceano Indiano", "ja": "インド洋", "ka": "ინდოეთის ოკეანე", "ko": "인도양", "ks": "ہِنٛدوستٲنؠ سۆدُر", "ku": "Okyanûsa Hîndê", "lfn": "Mar Indian", "mg": "Ranomasimbe Indiana", "mk": "Индиски Океан", "mr": "हिंद महासागर", "ms": "Lautan Hindi", "nan": "Ìn-tō͘-iûⁿ", "nl": "Indische Oceaan", "no": "Indiahavet", "nov": "Indiani Oseane", "nys": "Gabee-wodin", "pa": "ਹਿੰਦ ਮਹਾਂਸਾਗਰ", "pl": "Ocean Indyjski", "pt": "Oceano Índico", "ro": "Oceanul Indian", "ru": "Индийский океан", "sk": "Indický oceán", "sl": "Indijski ocean", "sr": "Индијски океан", "sv": "Indiska oceanen", "syc": "ܐܘܩܝܢܘܣ ܗܢܕܘܝܐ", "szl": "Indyjski Ôcean", "ta": "இந்திய பெருங்கடல்", "tr": "Hint Okyanusu", "uk": "Індійський океан", "ur": "بحر ہند", "uz": "Hind okeani", "vi": "Ấn Độ Dương", "vo": "Lindean", "wuu": "印度洋", "yue": "印度洋", "zh": "印度洋", "zh-Hans": "印度洋", "zh-Hant": "印度洋"}, "rules": [{"variant": "alternate", "language": "grc", "perspectives": "NULL", "value": "Ίνδος θάλασσα", "between": "NULL", "side": "NULL"}, {"variant": "alternate", "language": "mr", "perspectives": "NULL", "value": "भारतीय महासागर", "between": "NULL", "side": "NULL"}, {"variant": "alternate", "language": "uz", "perspectives": "NULL", "value": "Hind", "between": "NULL", "side": "NULL"}]}

source_tags

I include these because it's noteworthy to data consumers that source_tags is in fact not all of the tags. This is mildly surprising but completely rational from a data deduplication perspective. However, it looks like while the name:* tags are scrubbed, alt_name:* remain (despite the variant rules that also capture these). This isn't problematic; just inconsistent.

{"alt_name:grc": "Ίνδος θάλασσα", "alt_name:mr": "भारतीय महासागर", "alt_name:uz": "Hind", "gns:dsg": "OCN", "gns:uni": -2149804, "place": "ocean", "source": "GNS", "source:name:br": "ofis publik ar brezhoneg", "source:name:nys": "Derbal Nara website http://www.derbalnara.org.au used with permission", "sqkm": 74900000, "wikipedia": "en:Indian Ocean", "wikipedia:fr": "Océan Indien"}

[Dana Bauer (danabauer)]

Thanks, Ian Wagner (@ianthetechie), this is terrific feedback. I'll share it with the team.