The async_publish_time.cpp code has a bug: the connection is closed before starting to publish timestamp messages

[brocoli12138]

I found that the async_publish_time.cpp file cannot send timestamp messages correctly.
To find out why, I canceled the "send when dissconnected" connect option. It showed me that connection had been closed before publishment.
the code here is what I changed.
auto createOpts = mqtt::create_options_builder()
.send_while_disconnected(true, true)
.max_buffered_messages(MAX_BUFFERED_MESSAGES)
.delete_oldest_messages()
.finalize();

I change this part of code into this:

auto createOpts = mqtt::create_options_builder()
// .send_while_disconnected(true, true)
.max_buffered_messages(MAX_BUFFERED_MESSAGES)
.delete_oldest_messages()
.finalize();

Then I compile and execute this code, still can not receive message on server side.
This is part of code from "async_publish.cpp"

cout << "\nConnecting..." << endl;
mqtt::token_ptr conntok = client.connect(connOpts);
cout << "Waiting for the connection..." << endl;
conntok->wait();
cout << " ...OK" << endl;

and this is part of code from ”async_publish_time.cpp“：

cout << "Starting connection..." << endl;
cli.connect(connOpts);
auto top = mqtt::topic(cli, "data/time", QOS);
cout << "Publishing data..." << endl;

What's the difference? It is "->wait()"!!!! right!
After I corrected the ”async_publish_time.cpp“ into this:

cout << "Starting connection..." << endl;
cli.connect(connOpts)->wait();
auto top = mqtt::topic(cli, "data/time", QOS);
cout << "Publishing data..." << endl;

Then, I can publish data/time properly!

[fpagliughi]

Oh, yes, I see it. You're right. I will fix it.

[fpagliughi]

Oh, actually I did all that before I noticed my own comment...

        // Note that we start the connection, but don't wait for completion.
        // We configured to allow publishing before a successful connection.

This means I was specifically demonstrating a feature of the Paho C library where you should be allowed to publish messages before you're connected.

I need to look at this further.