You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In a pre 1.0 world, new features were added in patch releases and breaking changes in minor versions. Now, new features come in minor versions and breaking changes in major versions. The assumption underpinning this proposal is that you never care about running anything but the latest patch version of a particular minor branch.
For versions >=1.0 I would like to suggest:
We create (eg) dbt-snowflake-1.2 rather than dbt-snowflake.1.2.x
We use major/minor only in version files
Additionally, I propose a command to upgrade the patch versions of both the adaptor/core in an environment (in an ideal world an environment variable might control auto-upgrading patch versions if available).
I have been running something similar by manually maintaining it locally for the past few months:
When a new minor version comes out, bump my ~/.dbt/version file to 1.x.0
When a new patch version comes out, activate the relevant venv and pip install --upgrade dbt-core dbt-snowflake
Personally I find it works really well, and saves me from "environment sprawl" just to get the latest patch version. Hoping to solicit some opinions before looking at implementation options.
The text was updated successfully, but these errors were encountered:
In a pre 1.0 world, new features were added in patch releases and breaking changes in minor versions. Now, new features come in minor versions and breaking changes in major versions. The assumption underpinning this proposal is that you never care about running anything but the latest patch version of a particular minor branch.
For versions >=1.0 I would like to suggest:
dbt-snowflake-1.2
rather thandbt-snowflake.1.2.x
Additionally, I propose a command to upgrade the patch versions of both the adaptor/core in an environment (in an ideal world an environment variable might control auto-upgrading patch versions if available).
I have been running something similar by manually maintaining it locally for the past few months:
~/.dbt/version file to 1.x.0
pip install --upgrade dbt-core dbt-snowflake
Personally I find it works really well, and saves me from "environment sprawl" just to get the latest patch version. Hoping to solicit some opinions before looking at implementation options.
The text was updated successfully, but these errors were encountered: