Sorry to harp on this point that isn't fun performance engineering, but I think the tricky bit comes in exactly where you draw the line.
Is offering Postgres-as-a-service with PgDog under the hood okay?
What about a document-oriented, Firebase-equivalent that has PgDog and Postgres under the hood?
What about a low-code application development platform?
What about all of the above, but internal-facing vs paying-customer facing? Does it matter if the internal-facing is paid for via a cross-charge rate (e.g. a multinational company that charges divisions in other countries for services)?
Does a customer of all of the above services have to AGPL their code and release it publically?
Maybe it's all obvious, but that is the sort of thing I worry about with AGPL commercially. Even though I love OSS.
Is offering Postgres-as-a-service with PgDog under the hood okay?
What about a document-oriented, Firebase-equivalent that has PgDog and Postgres under the hood?
What about a low-code application development platform?
What about all of the above, but internal-facing vs paying-customer facing? Does it matter if the internal-facing is paid for via a cross-charge rate (e.g. a multinational company that charges divisions in other countries for services)?
Does a customer of all of the above services have to AGPL their code and release it publically?
Maybe it's all obvious, but that is the sort of thing I worry about with AGPL commercially. Even though I love OSS.