Feature flags can’t work for everything; a more complete testing methodology is required, and I hope there is more planned than we see here. One issue with fea

Feature flags can’t work for everything; a more complete testing methodology is required, and I hope there is more planned than we see here. One issue with feature flags is identifying what a “feature” even is. A single flag is insufficient to enable target features in systems of interconnected frameworks. There will be an entire graph of dependencies! Testers will need to understand the series of flags that need to be enabled for a target feature to work. And of course, there is a risk that this will make disabled features seem like bugs themselves! I am imagining Apple colleagues having to waste time reminding their testers how to enable a feature, after receiving a “bug” report about something not working.