Extract artifacts of previous stages in correct order
Overview
Currently, artifacts of previous stages are downloaded and extracted in random order (I assume that they run in parallel). This can be problematic if the same file (with different contents) exists in more than one artifact. What I expected was that the artifacts are extracted in order of their stages, so that the file of the latest stage overwrites the files of previous stages.
Therefore, I propose either to always follow the order of stages when extracting artifacts, or at least make it possible to be configured somehow (project-wise or in the yaml file).
Problem to solve
When a pipeline has multiple files with the same name from different jobs extracted out of order it can cause unexpected and unreliable results when extracted and used in a later job because order of loading the artifacts is undefined.