beckn / beckn/protocol-specifications-v2
No support for localizing authored content (multilingual name/shortDesc/similar fields)
- Dominant language
- No language data
- Stars
- 15
- Forks
- 13
- PR merge metrics
- No merged PRs in 30d
Description
**Background**
[Fabric-support #96](https://github.com/Networks-for-Humanity/fabric-support/issues/96) (ION) flagged that `Descriptor` only supports single-language strings for `name`/`shortDesc`/`longDesc`. Their workaround, `LocalisedText`, duplicates those fields rather than solving the problem at the attribute level.
The spec already has a documented position on localizing **terminology** — NFO-specific `context.jsonld`/`vocab.jsonld` aliasing into canonical IRIs (see [Schema Design Guide §"localize a Beckn network"](docs/Schema_Design_Guide.md#L614)). That mechanism fits enumerated/coded fields (statuses, categories, unit codes): the code stays a stable identifier, and its display label is resolved externally per language.
It does not cover **authored content** — free text written per-record, in more than one language, by whoever lists the entity (product names, descriptions, media labels). There's no external vocabulary to alias into here, because the content itself differs per instance, not just per language.
Contributor guide
Research direction
Start by reviewing the Descriptor fields and the Schema Design Guide section on localizing a Beckn network, then compare that terminology approach with the authored content fields named in the issue. Done requires a decided, documented specification approach for representing multilingual name, shortDesc, longDesc, and similar per-record content.
Written by the indexing model from the issue text.
Assessment
- Domain
- api
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100