In gh-313, it looks as though bbi truncated the error message, and @seth127 reports seeing the same. As a first step to hopefully avoiding that behavior, we should try to come up with a reproducer that shows this behavior.
(Note that it's possible that this truncation could be at a higher-level, such as bbr, or could depend on other details of the environment, like being executed from RStudio.)
In gh-313, it looks as though bbi truncated the error message, and @seth127 reports seeing the same. As a first step to hopefully avoiding that behavior, we should try to come up with a reproducer that shows this behavior.
(Note that it's possible that this truncation could be at a higher-level, such as bbr, or could depend on other details of the environment, like being executed from RStudio.)