Why wasn't this a good fit for database/sql?

[bradleypeabody]

Hello,

I realize this project is quite old, and perhaps the answer to this question is obvious to experienced users of Cassandra. In your explanation of the deprecation of this package you say it's not a good fit for database/sql - I'm just curious why. Tools like sqlx and modl work on top of database/sql and provide additional databinding functionality that is really useful (and take this burden off of the driver developers). I'm not an experienced Cassandra user, but from reading up on it (and from the fact that you were able to successfully make a working database/sql driver), it seems like this is a sensible approach.

If you could describe a bit more on your motivations for deprecating this package, that would be appreciated and I think would assist other people evaluating various solutions to determine the best course of action. I'm trying to decide between several different database solutions for an upcoming project (written in Go of course) and how the interaction with the database works will make a big difference.

Thanks for any info you can provide.

Best, Brad