Validation for make "strategy:depend" the default for parent-child pipelines
## Problem Statement <!-- What is the problem we hope to validate? Reference how to write a real customer problem statement at https://productcoalition.com/how-to-write-a-good-customer-problem-statement-a815f80189ba for guidance. --> The question we want to answer is whether users of parent-child pipelines want and expect the parent pipeline to always wait for child pipelines to complete first, so that we can make this a default behavior. Today not all parent-child pipelines are configured with the `strategy:depends` syntax which is a strategy that makes the parent pipeline wait for child pipelines to complete before the parent pipeline can complete. When a parent pipeline is configured to list all artifacts produced by itself and its descendant pipelines, the absence of the `strategy:depends` causes a race condition where the parent pipeline could complete before all the child pipelines have completed. The consequence is as soon as the parent pipeline completes, the MR reports shows an incomplete list of artifacts for those generated by the parent pipeline and by **some** of its child pipelines, but there are missing artifacts for child pipelines that are still running. Eventually all artifacts are visible in the MR as the remaining child pipelines complete, but until then the MR report is inaccurate due to incompleteness. ## Reach <!-- Please describe who suffers from this problem. Consider referring to our personas, which are described at https://about.gitlab.com/handbook/marketing/product-marketing/roles-personas/ --> <!-- Please also quantify the problem's reach using the following values, considering an aggregate across GitLab.com and self-managed:--> 6.0 = Impacts a large percentage (~50% to ~80%) of the above. ## Impact <!-- How do we positively impact the users above and GitLab's business by solving this problem? Please describe briefly, and provide a numerical assessment:--> 2.0 = High impact ## Confidence <!-- How do we know this is a problem? Please provide and link to any supporting information (e.g. data, customer verbatims) and use this basis to provide a numerical assessment on our confidence level in this problem's severity:--> 80% = Medium confidence ## Effort <!-- How much effort do we think it will be to solve this problem? Please include all counterparts (Product, UX, Engineering, etc) in your assessment and quantify the number of person-months needed to dedicate to the effort. For example, if the solution will take a product manager, designer, and engineer two weeks of effort - you may quantify this as 1.5 (based on 0.5 months x 3 people). --> 3 people x .5 months = 1.5 effort
issue