add article about generated C++ interfaces #33
Conversation
| Each setter methods returns the struct itself to enable (method chaining)[https://isocpp.org/wiki/faq/ctors#named-parameter-idiom}. | ||
|
|
||
| <div class="alert alert-warning" markdown="1"> | ||
| <b>TODO:</b> Need non-colliding naming |
There was a problem hiding this comment.
Using a set__ suffix seems like that most readable and non-colliding name to me.
There was a problem hiding this comment.
I did that already in 4d4b712 but I am not sure I like it.
dirk-thomas
force-pushed
the
generated_cpp_interfaces
branch
from
327b9e9
to
ec9a06a
Compare
dirk-thomas
added
in review
|
I have update the article based on the discussion in the meeting and removed all references to copy / move constructors / assignment operators. Please notice the change of the pointer types. @ros2/owners Please review. |
|
+1 |
1 similar comment
|
+1 |
| * `Ptr` and `ConstPtr` | ||
| * `SharedPtr` and `ConstSharedPtr` | ||
| * `UniquePtr` and `ConstUniquePtr` | ||
| * `WeakPtr` and `ConstWeakPtr` |
There was a problem hiding this comment.
Should we also add auto_ptr here? Or should we remove any of the above?
There was a problem hiding this comment.
I'd say no; auto_ptr is deprecated in C++11. The others look fine. People might find it weird that Ptr is now a "raw" pointer rather than a shared pointer, as it is in ROS 1. Not sure what else we could do though. Maybe RawPtr and ConstRawPtr, but I don't really like that either.
There was a problem hiding this comment.
-1 for RawPtr. The definitions in the document for Ptr types looks fine to me.
|
+1 |
dirk-thomas
force-pushed
the
generated_cpp_interfaces
branch
from
ec9a06a
to
0ed7369
Compare
add article about generated C++ interfaces
No description provided.