Meshtastic (Part 2)
As over a year has passed since I last wrote about Meshtastic, and as my ongoing effort with the technology has ceased, it seems like the right time to pick up from my last article.
Since I first purchased the hardware and got things operational, my development / usage of the platform has sadly been minimal. After developing a custom web interface to provide a friendly way to use the platform, and creating a messaging integration to allow me to send/receive messages securely while on the move, I had high hopes for using the platform as an alternate means of communication (hopefully with like-minded enthusiasts), however things didn't work out that way.
Despite trying different pieces of equipment, changing antennas, changing cables, verifying everything that could be verified, the sad truth is that from a human perspective it was a ghost town. While the debug logs showed a wealth of nodes sending their telemetry on a frequent basis, messages from human beings were seldom received. Worse, sending messages periodically to see if anyone was listening / would respond, was frequently met with silence.
After a few months (and many firmware revisions) I left it running to see how often people were messaging (or at least, their messages would reach me), and it was an incredibly small number. Following some of the published videos around using Meshtastic in the UK I occasionally switched between LONG_FAST and MEDIUM_FAST (in case everyone had moved over), however that made things progressively worse.
About a month ago (at time of writing) I decided to refresh my equipment and try one last time in case the situation had changed and the network was more lively. Sadly, the situation had worsened from where it was before. This time, my relay connection to London had gone which resulted in fewer nodes than before (and no coverage of the East). Whilst my connection to the North West had improved, messages from humans were still sporadic and replies never came. There are only so many times you can go up a tall tree to check everything is wired correctly before you realise you don't like ladders.
It was also at this point I discovered Meshcore (somewhat late to the party), and found many forums discussing how people had migrated from Meshtastic due to issues with messaging / the network being flooded with telemetry. While this sounded promising at first, studying the architecture of Meshcore made me realise it wasn't what I was after (and likely never would be).
Documenting what I learnt is somewhat tricky, as I never got to fully utilise the platform as I had hoped. The lack of height for my antenna definitely had an impact (even when bolted to the top of a large tree with line of sight for miles), and no cavity filter / cabling is going to fix that. The outdated nodes that have been abandoned / forgotten about still prove problematic for the network, being yet another issue that likely wont be fixed.
Overall I feel the platform could be very useful in certain scenarios, but for what I had hoped / envisaged, it wasn't to be. Who knows, maybe in a few years I might dig the bits out of storage and give it another try...