• Colloquially, server-side applications wait by the phone (at a particular extension), praying that someone—no, anyone!!—will call.
• Formally, server-side applications create a server socket that listens to a particular port.
  • The server socket is an integer identifier associated with a local IP address (think phone number) and port (think phone extension).
  • You should also think of the port number as a virtual process ID that the host associates with the actual pid of the server application.

• Full version of the server code (minus the createServerSocket interface and implementation) can be found right here.
• Our time server accepts all incoming connection requests and quickly publishes the time without so much as listening to a word that the client has to say.

static const short kDefaultPort = 12345;
static const int kWrongArgumentCount = 1;
static const int kServerStartFailure = 2;
int main(int argc, char *argv[]) {
  if (argc > 1) {
    cerr << "Usage: " << argv[0] << endl;
    return kWrongArgumentCount;
  }

  int serverSocket = createServerSocket(kDefaultPort);
  if (serverSocket == kServerSocketFailure) {
    cerr << "Error: Could not start server on port " << kDefaultPort << "." << endl;
    cerr << "Aborting... " << endl;
    return kServerStartFailure;
  }

  cout << "Server listening on port " << kDefaultPort << "." << endl;
  while (true) {
    int clientSocket = accept(serverSocket, NULL, NULL);
    publishTime(clientSocket);
  }

  return 0;
}

• The implementation itself, however, isn't the focus. Strictly speaking, it's generating dynamic content—uninteresting, but dynamic, nonetheless—and publishing it back to the client over the socket descriptor.

static void publishTime(int clientSocket) {
  time_t rawtime;
  time(&rawtime);
  struct tm *ptm = gmtime(&rawtime);
  char timeString[128]; // more than big enough                                                                                                                                           
  /* size_t len = */ strftime(timeString, sizeof(timeString), "%c\n", ptm);

  size_t numBytesWritten = 0, numBytesToWrite = strlen(timeString);
  while (numBytesWritten < numBytesToWrite) {
    numBytesWritten += write(clientSocket,
                             timeString + numBytesWritten,
                             numBytesToWrite - numBytesWritten);
  }
  close(clientSocket);
}

• The first five lines here produce the full time string that should be published. Let these five lines represent more generally the server-side computation needed for the service to produce output. Here's is the current time, but it could have been a static HTML page, a Google search result, an RSS XML document, an image, or a Netflix video.
• The remaining lines publish the time string—we'll call it the payload—to the client socker using the raw, low-level I/O we've seen before.