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
These are the the high level proposed changes before we start working on them
Remove docker support but only for people running on their local machines, refactor the docker files to a subfolder to keep gitpod support
From the discussion - we should not support docker as it confuses the setup further for most people especially on a windows setup. A lot of people showing up for the interview have good knowledge of how to setup Spark on their machines. One outcome of this is that
Remove batect for running tests
Not many people are familiar with batect and find it hard to understand their way around it, replace it completely with pytests
Code refactoring
src/it - this should be ideally in src/test/it which is the more standard convention used for Scala
Remove geo location from package names and rename org organization := "com.thoughtworks.cd.de" to organization := "[com.thoughtworks]",
For the scala version - remove the use of any curried functions and implicit classes. For python simplify the code further and reduce method bloat.
This is just the general set of things, lets add more things to the list and we can start merging these changes one by one. This is in parallel to the changes being planned to support databricks. Please add your thoughts. @lauris-tw@lleites@darshanj@jmolina4@svishal9
The text was updated successfully, but these errors were encountered:
Docker/Gitpod: Different set of people have different levels of familiarity with the code. I feel we should be open to more ways to interact with code and hence keep docker and other ways while adding others.
While we are talking about adding ways, it would be great to add CI/CD as well for executing tests.
Batect: This was more of a dog food. We were contributing to batect and it was a chance to use the simplification which it offers. Having said that, more than happy to support combination of pytest & pyspark test if it is adding complexity.
Code refactoring: +1 to all points. We might also look at some of the tech debts.
These are the the high level proposed changes before we start working on them
Remove docker support but only for people running on their local machines, refactor the docker files to a subfolder to keep gitpod support
From the discussion - we should not support docker as it confuses the setup further for most people especially on a windows setup. A lot of people showing up for the interview have good knowledge of how to setup Spark on their machines. One outcome of this is that
Remove batect for running tests
Not many people are familiar with batect and find it hard to understand their way around it, replace it completely with pytests
Code refactoring
This is just the general set of things, lets add more things to the list and we can start merging these changes one by one. This is in parallel to the changes being planned to support databricks. Please add your thoughts. @lauris-tw @lleites @darshanj @jmolina4 @svishal9
The text was updated successfully, but these errors were encountered: