Meshtastic - “The reports of my death are greatly exaggerated.”

Many of you will have discovered my cases via Andy Kirby on YouTube and may have seen a video he posted on 14th July 2024 proclaiming “Meshtastic is broken…”. That title/introduction “ruffled the feathers” of many in the community together with businesses invested in Meshtastic, but it is not "pining for the Fjords" as some have jumped to conclude.

I know Andy loves a clickbait title and thats exactly what it was, however the initial intro and sentiment was harsh and Andy has held his hands up and acknowledged that in a subsequent video (more on that video in a second) However… we have to look at where that sentiment comes from.

In my discussions with Andy he said that he just tries to represent the experiences of the community which he mainly see’s on Discord (where a lot of UK based Meshtastic users gather, many areas of the UK having a dense mesh) The issue I see is those with the loudest voice are always those with a complaint, those without complaint being rarely heard. It’s therefore likely Andy is always going to see more negative comments than positive, but again we have to look at what those negatives exactly are because they exist.

“Messaging doesn’t work” is the loudest complaint. “No other nodes in my area” is another common one. Both of these likely represent the two extremes of the “bell graph” representing the density of the Mesh around you. At one end you have a dense mesh with many nodes, at the other end a mesh lacking nodes. If the mesh around you has many nodes you may experience congestion which causes messaging to be unreliable, while a lack of nodes around you is a lonely place!

The latter is easy to answer, “give it time” and see if you can co-ordinate building a mesh with others, thats fun! If the former is the problem though what’s the solution? Meshtastic themselves published a suggestion for densely populated meshes which is switching the LoRA Data rate of the nodes to “Medium / Fast” versus the default “Long / Fast”. The logic here, I believe (remember not a radio nerd), is that the airtime needed to transmit a message is much shorter therefore decreasing congestion. The price though is a reduced range but in a dense mesh that may not be an issue.

Andy’s most recent video is pushing that change in the UK where we have areas of very dense mesh (my node list was 200+) I think for areas with a dense mesh will be an interesting experiment and I have switched to see what happens in the coming weeks.

The difficulty of course is getting the nodes to all change to medium fast…. This highlights the other challenge with Meshtastic which is co-ordinating the set-up of the Mesh (this is in part why I’m writing this) It’s a good idea to join a Discord community or FaceBook group to discuss the setup of the mesh in your area.

Still I’m left wondering is congestion actually the issue though? There have been suggestions that MQTT can cause issues (a recent software update is targeting this I believe), devices with old firmware perhaps can be an issue and nodes with incorrectly set roles. The roles suggestion is actually worth considering, if you have a primary node at say a high point in your house and then carry another with you around the house it’s recommended that the node you carry be set to “CLIENT_MUTE”. That node will then not relay messages, something that is pointless as your node in the high point will already be relaying them with better success.

I think the truth is no one has a definitive answer to why messaging can be unreliable in a dense mesh, congestion is not the only factor. All I’d say is Meshtastic is new technology, still very much in development and whats key is it has exploded in growth. What we are seeing is the very definition of “teething issues”. Think of Meshtastic as a new baby that is growing, it has amazing potential but it’s only crawling right now will one day walk and run. In the mean time we have to have patience, experiment and co-ordinate in order to learn how to help it walk and run.

That is ultimately why I bought a HeltecV3 and made Bender in the first place, I wanted to learn and be part of creating something that will be genuinely useful.