I stopped using PainlessMesh and started writing this mesh based on ESP-Now quite some time back. I've never quite finished it, but ESP-Now broadcast frames allow you to get things in pretty close sync probably in the 10s of ms range.

@@NickReynolds Is there any way I can get some insight in how you wrote it? Could i parallelise the message sending with the sync elements so I won't get a problem if I have a transmission at the same time of a sync?

I'm afraid I've never quite finished it. I've started looking at it again recently.

@@NickReynolds im more than happy to help you with this in a few months!

@nick - rookie here again, what would you recommend now that two years have gone by if I guy wants to replicate this kind of ninja for his portable esp32 cam pit motion detector network for securing up air bnbs.

Yes, my intention is to release this once it is in a fit state. I am making slow progress towards this.

Nick Reynolds any update on releasing it? I’ve been using painlessMesh but would be interested in trying a ESP Now version.

I am still slowly working on this, but it's a hobby project and quite time-consuming.

I'm not actually using broadcasts, for things that need to go to all peers it's unicast to them all.

Each node uses the concept of 'echo quality', ie. does it see its own packets repeated back to it. Poor echo quality roughly implies weak signal between two nodes.

I have implemented my own mesh network using ESP-Now as the transport and a variant of the routing algorithm from BATMAN. It is a variant of it as BATMAN doesn't map directly onto the features available in ESP-Now, relying on broadcasts. www.open-mesh.org/projects/batman-adv/wiki/BATMAN_IV

@@NickReynolds woow. thks so much for the info. I'm working with espmesh but you are far away from my knowledge.

@@danielbazzi4325 I'm just at the first stages of getting this working. Once it reaches a stage where it might be usable by others I plan to release it as a library. The reason I am doing this is that I found the mesh solutions that use conventional AP+STA mode worked but responded badly to packet loss or changes in topology if things moved and my project will involve wearable devices. It also MIGHT mean longer range between nodes but that's hard to quantify.

A much later version of this is now on GitHub. It needs further work, I could really do with constructive feedback on it. github.com/ncmreynolds/m2mMesh

@@NickReynolds I apprectiate that thank you, I'm a novice dodder so I shall do my best and give some feedback.

No, this is me writing my own Arduino library to do this.

@@NickReynolds can u please make tutorial on this if this is okay for you .

I've been doing a bit of work in the background but I'm not yet happy with the code so haven't released it. The project I need it for is in March so I'll be concentrating on it in February.

I'm afraid I've never quite finished it. I've started looking at it again recently.

It's not specifically limited, but eventually you'll run out of heap and right now I'm using uint8_t as an index that would limit things at 254. There's no reason that uint8_t couldn't be an uint16_t but I bet you'd have run out of heap by then anyway. I've only managed to get it to 16 nodes but that's because I ran out of ways to power the ESPs. As I'm using ESP-Now you can have 20 peers per node, but that's just local peers and the mesh as a whole can be bigger. NB: peer to peer communication is NOT encrypted, this is not a requirement for what I'm doing. I've done some artificial capping of local peers to test managing that but really need to build more of a test rig to get real world answers. I've got a drawer full of ESP-01s kicking around unused, just need to solder up a way to plug lots in at once. At the moment I'm doing routing advertisements more frequently than is really necessary and this will limit performance too. This is me feeling my way with a work in progress, not a delivered project.

Since then I've done a massive rewrite during pandemic lockdown, as modern versions of the Arduino ESP8266 core support ESP-Now broadcasts. It handles discovery much better and scales better, but I still haven't got around to a version I'm happy to release.

No, they are electrical junction boxes.

These are Wemos D1 mini pro, because I like the D1 boards for prototyping and they have an external antenna.

Your comment has made me come back to this and try to knock it into acceptable shape. While I've been using it myself a bit, it's not consistent or documented and lacks examples. Trying to get a few of the essential info fixed up at the moment.

@@NickReynolds Good. You can email me and we can see if I can help you on that.

@@NickReynolds Hi Nick. Any progress on this? Please contact me, i will help also on this..

​@@klausjuhantalo8415 I've been doing loads of work over the last weeks trying to get the code I removed during a big rewrite integrated back in, a full set of examples and some basic documentation sorted. You don't seem to have any contact details visible.

@@NickReynolds Now i have, ukposts.info/the/yg2CEGvBOTz2kMykC3t78Q.htmlabout

This is done with a mesh network library I've been working on sporadically but it's not stable enough for release just yet. I'm doing some work on it over the next few weeks so it might get there. I keep saying that though.

This is a very old version of what is now m2mMesh github.com/ncmreynolds/m2mMesh

I'm still working very slowly on making the code solid enough to release. It works but doesn't scale as well as I would like. I've had about 30 nodes working but it starts to break down and things drop out of the mesh at that point. I'm doing a minor rewrite where it does much less spamming of the network with updates.