Khyber Sen

The error from `tests/statics` (the above is in `tests/structs`) is from this test failing: https://github.com/immunant/c2rust/blob/e42598e1ca01161b8921096b53590ad538e127db/tests/statics/src/test_sections.rs#L32-L42

> I think there should be some limits on that policy. For example, all of `u8..usize` and `i8..isize` types are technically just identifiers which may be shadowed in user code,...

> Adding `impl Make for T` for various `N`: I'm not sure this is worth the concision it enables; inferring which kind of syntax node we produce is not the...

Okay, I haven't really looked into the details of what `immunant/c2rust-testsuite` does, so that might make sense. I do think running the in-repo `./scripts/test_translator.py tests/` as part of `cargo test`...

By the way, the issue with `cargo run -- transpile` not rebuilding `c2rust-transpile` and having to do `cargo build && cargo run` is really annoying and not easy to figure...

Could we instead wrap a Rust cbor crate with C bindings and use that from `c2rust-ast-exporter`? That way `cargo` handles all the dependencies and we use the same codebase for...

We just use these functions: ``` cbor_encode_boolean cbor_encode_byte_string cbor_encode_double cbor_encode_int cbor_encode_null cbor_encoder_close_container cbor_encoder_create_array cbor_encoder_get_buffer_size cbor_encoder_get_extra_bytes_needed cbor_encoder_init cbor_encode_string cbor_encode_string_array cbor_encode_text_string cbor_encode_text_stringz cbor_encode_uint ``` and [`minicbor::encode::Encode`](https://docs.rs/minicbor/latest/minicbor/encode/struct.Encoder.html) has all the equivalent methods.

Making `tinycbor` a git submodule might work (https://github.com/immunant/c2rust/issues/470#issuecomment-1172257062). But I don't think creating C bindings for `minicbor`, specifically the type-specific ones in [`minicbor::encode::Encode`](https://docs.rs/minicbor/latest/minicbor/encode/struct.Encoder.html) to match exactly the ones we use...

It's off by default now and enabled with `--set-runtime`. I still think this is good for testing especially, since it ensures the runtime crate is up-to-date and points to the...

`lighttpd` is still using old-style K&R function definitions? Does that have anything to do with the error, i.e. the function being untyped?