Discussion for: Unicode 17 Support

[rkajiyama]

This proposal suggests adding a new Unicode 17.0 based collation family to MySQL: utf8mb4_1700_*.

Related bug report in Bug Databaes: Bug #120486 | Proposal for Unicode 17 Support

MySQL currently provides Unicode-versioned utf8mb4 collations based on UCA 9.0, such as utf8mb4_0900_*. Since then, Unicode has added many new scripts, CJK ideographs, symbols, emoji, and updated collation data. Applications that store modern multilingual text increasingly need SQL-level comparison, sorting, equality, and indexing behavior that matches current Unicode data.

The proposal is to add utf8mb4_1700_* as a new, explicit collation family rather than changing the existing utf8mb4_0900_* collations. This keeps current deployments stable while giving new schemas and opt-in migrations access to Unicode 17.0 behavior.

Without Unicode 17.0 collations, applications may need to rely on client-side comparison logic, which can diverge from MySQL behavior for ORDER BY, GROUP BY, DISTINCT, UNIQUE constraints, and indexed lookups. A versioned utf8mb4_1700_* family would keep MySQL’s string comparison model current while preserving backward compatibility.

[villagedominic]

This created a major issue for us. You can see the details here -

Issue - villagesql/villagesql-server#353
Issue - villagesql/villagesql-server#811

We would like to see the encoding made consistent. We are also supportive of adding more optional encodings, as mentioned.