Closed RobMcZag closed 4 years ago
Hi, Thanks very much for bringing this to our attention. At the moment we are only officially supporting Snowflake, with BigQuery support in the works. We have fully tested dbtvault on Snowflake, but no other platforms at the current time, so dbtvault may produce unexpected results on other platforms (such as Postgres)
I think fixes like this will likely be implemented along with development of full support for Postgres, which will likely be soon after we have finished porting to BigQuery.
I am currently working on a contributor/public development environment which will allow contributors to develop code using our test framework and ensure our test suite is run against Pull Requests. This will mean that future contributions like yours can be tested for quality and be merged into the package with confidence.
I will leave this open as a reminder for us when we get to implementing Postgres support, and when we have a more robust contribution pipeline in place. Both of these are on the horizon, so please stay tuned and thanks for contributing!
Hi Alex, I am happy you want to have a robust contribution pipeline.
It is a tiny change, should give no problems on any platform, but I agree it should go through all tests to avoid regressions.
I have the fix in my local version, so no hurry from me. Good Sunday evening !
Hi @RobMcZag In dbtvault v0.7.0 We've moved all subqueries to CTEs. This should have solved this issue as a result. Can you check it out and close this PR if all is well? Thanks
This was preventing the macro to work on Postgres.