V2 roadmap (post-V1 feature backlog)

[ihistand]

Umbrella for the post-V1 backlog. Full context + per-item notes live in SQLAnvil/docs v2_roadmap.md. Split an item into its own (sub-)issue when it's picked up.

• MySQL / MariaDB adapter (committed — do first)
• Import/export at the warehouse edges — files & object storage (GCS/S3/Azure CSV/Parquet/JSON, incl. BigQuery's native EXPORT DATA/external tables) and cross-platform DB↔DB movement, sharing one type-translation layer. V2 goal: a Postgres/Supabase equivalent of BigQuery's native object-storage read+write. Consumer file-sync (Dropbox/Drive) is a separate fetch-connector, out of scope. (See v2_roadmap.md §2.)
• Dataform → SQLAnvil conversion tool
• Interactive sqlanvil init (Q&A scaffolding)
• --environment (named dev/staging/prod environments)
• Pre-run SQL validation — sqlanvil validate / a real Postgres --dry-run (EXPLAIN-validate the DAG; needs a shadow schema for forward refs)