They shouldn't (all come from same source AFAIK) but could be some sort of explanation
its tricky with cheerligh stuff at moment as since the Twitter API changes - the main listener on thingsspeak is losing messages so trying to keep track of what;s going on is proving problematical
So throwing this into the mix has really not helped
I've always believed that gremlins do inhabit computing devices - I now believe that they are using Bell Continuum quantum signalling to co-ordinate actions!
I have difficulty seeing how that would cause the problem you show. If it is happening regularly you could just make the change node before it clear the topic.
Agreed but when something that's been working fine for months and you honestly believe that you haven't changed anything starts intermittantly failing then you start clutching at straws
I'm thinking about implications of not actually having it in the flow to simplfy it -I've always had it in to prevent duplicate mqtt messages going out but it might not have any real effect if I let them thru
If the messages are going to an mqtt node then different topics would be going to different destinations on mqtt anyway, so the node would be working as expected. If you think that the topics should all be the same, but you are seeing a regular apparent problem then put a switch node in, checking for the expected topic and sending to o/p 1, with Otherwise going to o/p 2. Then see if you get any messages out of o/p 2.