https://assets.woodwar.com/meshcore_sf_test_report.pdf
@sesivany
Thanks, that's really good information!
I was looking explicitly for reasons why you did not start with enforcing regions/scopes first.
Chapter 4 is dedicated to it. Great!
With this info there I'm wondering, wouldn't it require the very same amount of coordination to enforce regions first?
It would be easier to measure the impact because everything else stays compatible and the only "force" is the soft social pressure to use scopes to continue using the mesh, and setting the regions is indicated even with "broad support".
By the current move you cut off large parts of traffic because you disconnect from the rest of Europe and not every repeater can be switched over due to real-live reasons.
So it will definitely *feel* like a success first but you can't "prove" it - in contrast to enforcing regions where all parties are kept in the game.
And if *that* didn't work out, *then* try other radio settings.
Hmmmm...
Anyway, good luck. ♥️
@voyagermesh @jwildeboer I'm not from the Dutch community, I'm from the Czech community and we switched to a similar preset back in January where there were no regions and scopes. We did like a week-long trial of both presets and EU/UK one felt much slower and we also anticipated the problems the Dutch report now found. The community voted for faster and more scaling preset by a big margin.
IMHO mesh has scaling issues. Regions and scopes help, but if we want it to scale, especially if we want to use a single preset for the whole Europe and allow some interstate traffic, we have to use all means available.
@jwildeboer @sesivany @voyagermesh
Folks, you are replaying https://en.wikipedia.org/wiki/Packet_radio and APRS https://www.aprs.org/fix14439.html https://w6hs.net/using-aprs-part-one-digipeaters-and-path-settings/ history.
Any mesh based on slow and BW limited medium will eventually saturate.
Meshcore already faces issues with 100s of users. Can you imagine thousands?
You need a beefier https://en.wikipedia.org/wiki/Backhaul_(telecommunications)
That is why the modern HAMNet uses wifi (vs packet).
Or adopt high latency: https://en.wikipedia.org/wiki/Delay-tolerant_networking
Fiddling with presets will not help longterm.
@marsik
The presets will just hit another wall, I fully agree.
However I don't think we can switch to a different network mechanism like in your links, it now *is* a mesh on LoRa.
IMHO we should focus on the localization of traffic -- which is what regions/scopes are for.
This is not just a technical solution but also requires us to agree on common rules - in our collapse prepping world view that's a necessary society skill anyways.
Of course, it boils down to the same problems we fight for centuries: bad actors, ignorance, the tragedy of the commons, etc etc. but ultimately we *NEED* to stop trying to solve society problems, even technical society ones, with technology.
And IMHO regions/scopes are a good enough tool to divide and conquer the problem. You can organize locally, and apply similar rules on bigger scope.
Then again, it might be just wishful thinking and we struggle exactly because we can't generally resolve human misbehaviour.
This being said, I think that "Backhaul" principle is somehow applicable to mesh networks, even based on the existing region/scope mechanism as the localization mechanism.
Not that I have deeply thought about it but I don't wanted to sound too pushy against it in my post before.
@voyagermesh @jwildeboer @sesivany Oh, I was not proposing to move to APRS. Just that you learn from the dead ends there.
We can probably use the existing region API and router rules to emulate this.
Lets say that only when you mark your message with #wide, the top of the mountain repeater will repeat it.
Or we create two mesh frequencies, one for cross-county net and one internal. And "route" only non local messages to the backbone.
The zones might work too, but depend on geography.