Reader feature types that aren't connected to any other objects in your workspace are not read, therefore you don't need to worry about their performance impacts.

• A. True
• B. False
• C. It depends
• D. Not applicable

Answer: (B)

In FME, data flow is driven by readers feeding transformers feeding writers. Even if a reader’s feature type isn’t connected to any downstream objects, the workspace run can still involve that reader opening its data source to determine structure, metadata, or perform internal checks. That means there can be I/O and parsing overhead regardless of whether any features actually end up being processed downstream. So assuming that unconnected reader feature types aren’t read and won’t impact performance isn’t reliable. You can still experience startup delays, memory use, and I/O costs from those readers, especially with large datasets. To avoid unnecessary work, you’d typically disable or remove unused readers or gate their execution with parameters.