Is your feature request related to a problem? Please describe.
13 adds implementations of cryptographic operations based on the openssl crate, but using a C library like OpenSSL has both security implications and complicates cross-compilation.
Describe the solution you'd like
An alternative backend based on the RustCrypto library suite should be added.
Describe alternatives you've considered
There are numerous alternative libraries for the required cryptographic algorithms (or at least a subset of them) available (MbedTLS, wolfSSL, ...).
Adding backends for those might be worth exploring in the future, especially when it comes to embedded development.
For instance, MbedTLS is integrated into the ESP-IDF for the ESP32 and would provide hardware acceleration that RustCrypto currently does not.
However, RustCrypto has the advantage of being a collection of native rust crates that should work everywhere that dcaf-rs can be compiled without hassle, and should therefore be the higher priority.
Is your feature request related to a problem? Please describe.
13 adds implementations of cryptographic operations based on the
openssl
crate, but using a C library like OpenSSL has both security implications and complicates cross-compilation.Describe the solution you'd like An alternative backend based on the RustCrypto library suite should be added.
Describe alternatives you've considered There are numerous alternative libraries for the required cryptographic algorithms (or at least a subset of them) available (MbedTLS, wolfSSL, ...). Adding backends for those might be worth exploring in the future, especially when it comes to embedded development. For instance, MbedTLS is integrated into the ESP-IDF for the ESP32 and would provide hardware acceleration that RustCrypto currently does not.
However, RustCrypto has the advantage of being a collection of native rust crates that should work everywhere that
dcaf-rs
can be compiled without hassle, and should therefore be the higher priority.Additional context
None