Great idea, kind of a chaotic mess in practice. Better than nothing by far, but the industry I think will be eager to receive an improved alternative. The prob

Great idea, kind of a chaotic mess in practice. Better than nothing by far, but the industry I think will be eager to receive an improved alternative. The problem with any tool like Dbt that abstracts over differences in databases is that a huge amount of work goes into building "adapters" to support the various details and quirks of each supported database. That ends up being a substantial technological moat which inhibits the growth of competitor systems. Another option is to do what Datasette did and focus on supporting one specific database, gradually expanding to a second database after years of demand for it.