In the IP world with UDP it's up to the application to respond to failures. I use BLLogMonitor (awesome) to deal with devices like my door locks that often fail to respond. However, it's not really reasonable to think of building events for every device in you home to do the same (I realize many hardly/rarely ever give issue), but it seems like if this was built into the application logic itself it would make any system running it more stable/dependable, likely react faster to failures, and simply take less CPU time slices? Though I doubt the latter is that big of deal. I guess it could be if you actually had events to respond to every device you had failing.
Has this been considered? Maybe with a user defined number of retries...
Has this been considered? Maybe with a user defined number of retries...