So, I noticed that there are two very similar ways to break up channels when using the message service. I could not find a discussion of the pros and cons of each. Any thoughts?
It seems to me the only real difference is that with subtopics, the producer creates a separate stream of information, where as with using selectors, there is only one stream of data from which consumers pick out the bits they want. Are the methods equal from a network traffic stand point? Are there any differences from a memory management perspective? What are some situations where one would be preferred over the other?