We decided that for development & debugging we can use the log crate. @alexandruag raised some concerns with using log directly for production use cases. A problem with directly using log macros is that the attributed severity (i.e. error, warning or info) depends on the use case of the consumer. What is an error for one use case might be the just an info for another.
This is a topic that has been discussed during the rust-vmm sync meeting on 2021-03-08 (meeting agenda: https://etherpad.opendev.org/p/rust-vmm-sync-2021).
We decided that for development & debugging we can use the
logcrate. @alexandruag raised some concerns with using log directly for production use cases. A problem with directly using log macros is that the attributed severity (i.e. error, warning or info) depends on the use case of the consumer. What is an error for one use case might be the just an info for another.