Several things learnt from recent preparation of data for testing. Question the way when it takes too long to process data sed: remove last n lines In the past I always thought sed is the fastest way to manipulate data. Is it true for all the cases? I generated 1T testing data in ndjson format, however the last lines of those files are corrupted json data. I had to remove the last line.

The journey I worked on a dapr microservice project which need to read a list from a file and do some useful things. Argo-workflow supports loop over a list of items, however one of my steps always got "invalid character 'G' looking for beginning of value" before going to next step in the workflow template. The step is as following: - name: get-deps inputs: parameters: - name: changes script: image: ubuntu:latest command: [bash] #args: ["echo result was: {{inputs.