294 ios backend framework #299
Conversation
| #RENAME libtinyxml2.dylib | ||
| ) | ||
| endif() | ||
| if(NOT IOS) |
There was a problem hiding this comment.
You could just write if (APPLE AND NOT IOS), I think.
There was a problem hiding this comment.
Nope, because there is an else() at the end of the tinyxml2 block, that would then take linux and iOS :)
| BUILD_WITH_INSTALL_RPATH TRUE | ||
| INSTALL_NAME_DIR @rpath | ||
| MACOSX_FRAMEWORK_IDENTIFIER io.dronecore.backend | ||
| VERSION 0.0.1 |
There was a problem hiding this comment.
We can get the version later from the latest git tag.
grpc/server/src/backend.cpp
Outdated
|
|
||
| dronecore::DroneCore _dc; | ||
| dronecore::backend::ConnectionInitiator<dronecore::DroneCore> _connection_initiator; | ||
| std::unique_ptr<grpc::Server> server; |
There was a problem hiding this comment.
argh. I lost it in the rebase. And I actually verified that it was still _server xD. Will change, sorry.
There was a problem hiding this comment.
Oh, I actually messed up, because it fails the test. Just the CI is not compiling grpc for now :D.
grpc/server/src/backend.h
Outdated
| DroneCoreBackend(); | ||
| ~DroneCoreBackend(); | ||
|
|
||
| DroneCoreBackend(DroneCoreBackend &&op) noexcept; |
There was a problem hiding this comment.
Why is the noexcept needed? Shouldn't everything be compiled without exception anyway?
There was a problem hiding this comment.
Not, sure, I copied a pImpl example to not mess it up. Should I remove it?
There was a problem hiding this comment.
I guess so, not sure. Probably not important.
JonasVautherin
force-pushed
the
294-ios-backend-framework
branch
from
8e96747
to
cba697d
Compare
JonasVautherin
force-pushed
the
294-ios-backend-framework
branch
from
cba697d
to
8c567c7
Compare
|
|
||
| DroneCoreBackend::DroneCoreBackend() : _impl(new Impl()) {} | ||
| DroneCoreBackend::~DroneCoreBackend() = default; | ||
| DroneCoreBackend::DroneCoreBackend(DroneCoreBackend &&) = default; |
There was a problem hiding this comment.
What does default mean for these? What about using delete ?
There was a problem hiding this comment.
This time, I think it is explained in the article showing the pImpl coming from Scott Meyers' book (article here. Specifically, it says:
// Tell the compiler to generate default special members which utilize
// the power of std::unique_ptr.
// We can do it here because the implementation class is defined at this point thus
// std::unique_ptr can properly handle the implementation pointer.
There was a problem hiding this comment.
I can still change it later if you want, but as far as I understand, default is not wrong. So I'll already merge that :-).
There was a problem hiding this comment.
Let's continue this discussion here.
When building for iOS, the backend is now automatically built as an iOS Framework that can be directly added in an Xcode project and used (it only exposes the one
runBackend(port)function).There are two small glitches regarding dependencies:
-DBUILD_STATIC_LIBS) that we may not want to set for the whole project. So I'd like to build tinyxml2 separately (like grpc and others). For now, I don't link it in iOS.Fixes #294.