Is your feature request related to a problem? Please describe.
RpcClient is a good fit for one-off RPCs. However for a one-off use one has to remember to close the RpcClient otherwise, the channel will retain the consumer forever and when the channel is used some time later for another one-off RPC, the new and old consumer will collide.
It sure would be nice to have some support with not-forgetting this. I for one certainly need it.
Describe the solution you'd like
The preferred way to remind developers to call the close method surely would be to make RpcClient implement Closeable or at least AutoCloseable if we truly do not want the close method to be idempotent (I cannot imagine why we would want that, but maybe there's a reason it's written as non-idempotent now).
Describe alternatives you've considered
Implementing the try-with-resources-construct I want to use by writing the appropriate try-finally-block myself. This is a. ugly and b. error-prone, because it's one of the cases where one needs an extra layer of try-catch inside the finally-block. Of course I can look up what the correct pattern is, but we force that onto users?
acogoluegnes
commented
1 year ago
Is your feature request related to a problem? Please describe.
RpcClientis a good fit for one-off RPCs. However for a one-off use one has to remember to close theRpcClientotherwise, the channel will retain the consumer forever and when the channel is used some time later for another one-off RPC, the new and old consumer will collide.It sure would be nice to have some support with not-forgetting this. I for one certainly need it.
Describe the solution you'd like
The preferred way to remind developers to call the
closemethod surely would be to makeRpcClientimplementCloseableor at leastAutoCloseableif we truly do not want theclosemethod to be idempotent (I cannot imagine why we would want that, but maybe there's a reason it's written as non-idempotent now).Describe alternatives you've considered
Implementing the try-with-resources-construct I want to use by writing the appropriate try-finally-block myself. This is a. ugly and b. error-prone, because it's one of the cases where one needs an extra layer of try-catch inside the finally-block. Of course I can look up what the correct pattern is, but we force that onto users?
Additional context
No response