As noted by @notactuallyfinn in codemeta/codemeta#500 , the v1 and v2 contexts cannot be fetched by JSON-LD processors because of the DOI's content negotiation issues.
The workaround is for software processing JSON-LD documents to have a cached copy of the context (which they should regardless, it's a good practice both for reliability and reproducibility). We should document this.
As noted by @notactuallyfinn in codemeta/codemeta#500 , the v1 and v2 contexts cannot be fetched by JSON-LD processors because of the DOI's content negotiation issues.
The workaround is for software processing JSON-LD documents to have a cached copy of the context (which they should regardless, it's a good practice both for reliability and reproducibility). We should document this.