If for any reason, our definition of gpr_int64 is different from the definition of google::protobuf::int64, the virtual override would fail, as this needs to be of the precisely exactly the same type.
That change allows us to use protobuf's int64 definition.
tag. This can happen if the tag is actually an integer being
typecast to void*
To avoid breaking the API of existing Next calls, I've made
a new AsyncNext method with a tri-state return that indicates
whether there is a shutdown, an actual event, or a timeout.
Still needs proper testing for the AsyncNext method specifically.
The mappings are a combination of useless, inaccurate, and harmful.
Leave them out until they are helpful and correct.
Fundamental issues:
- If in need of mappings, a user would need HTTP -> gRPC code mappings.
But the docs don't provide that since the mappings are not 1:1.
- There _is no_ gRPC -> HTTP code mapping taking place in gRPC. This
may change in the future, but until then the docs are misleading.
But even given those were resolved, there are additional issues. Two
obvious examples:
- 501 Not Implemented is refering to HTTP methods such as GET and POST
not being implement. That should not be used for gRPC methods not
found (UNIMPLEMENTED).
- 404 Not Found is for when the resource is not found. In gRPC, the
resource is the gRPC method, not the parameters (NOT_FOUND). If there
is a REST conversion layer on top of gRPC, then 404 would be
appropriate, but that just shows that the mapping can not be
canonical.
- Renaming default credentials -> google default credentials.
- Various other things in cpp:
- Adding Cpp wrapping for JWT Tokens.
- Renaming ComposeCredentials -> CompositeCredentials.
- Remove CredentialsFactory as it's unnecessary
- Effectively make Credentials a channel factory, allowing different credential types to create different channel types - this gives us a hook so that InsecureCredentials can at runtime instantiate a different kind of channel as required - giving us a way of generating an openssl free version of grpc++.
- Server credentials not touched yet, but they'll need to be updated.
- Tested with new tool (print_default_creds_token) on:
- workstation for env var and well known place.
- GCE for compute engine default creds.
- I'd prefer the grpc_default_credentials_create() API to remain
synchronous even though there may be an async call for gce detection
on which we block.