HttpClient implementation that uses a cache; we put a timeout around looking up a cache entry, but when this returns on the error channel (which includes there's no cache entry), we have a Fiber terminated with an unhandled error logged where the cause is our NoCachedResponseFound error. However, the pipe remains on the error channel, where we then handle the error. So it behaves as we expect, except for the concerning log message. Is there likely to be anything strange about using a timeout inside a HttpClient?