Chat freely about anything...

User avatar
By sfranzyshen
#57410
rudy wrote:I had some exceptions and I tried to manually create a directory for the exception tool to use. I don't know why it still didn't work. I followed the information provided here and verified it from the issue raised on the Github page. I ended up messing something up with my environment and I can't get the source to compile. I tried uninstalling and deleting everything and starting fresh and it still isn't working. I don't know how long it will take to get back up and running.

ouch? ... sorry
User avatar
By sfranzyshen
#57411
rudy wrote:I have been running with seven ESPs and it has been working pretty good. I have serial terminals open so I can watch the received packets. I have a serial routine that I use to inject my own packets in addition to the example's communication test. I slowed down the rate of transmissions that the example had. It was impossible to watch with so many nodes. Now I'm typically getting a message in every 1-5 seconds. The odd time closer.

0x8 jsonSubConnCount(): leaving count=1
0x8 jsonSubConnCount(): leaving count=1
0x8 jsonSubConnCount(): leaving count=1
0x8 jsonSubConnCount(): leaving count=1
0x8 jsonSubConnCount(): leaving count=1
0x8 jsonSubConnCount(): leaving count=1

Added the code you suggested but all that shows are thousands of lines of the above that obscure the issues.


I totally agree with you on this. I can't deal with all that. But is likely to be where most of the answers will be found. So I'm going to look at it again and see if I can eliminate the noise.


It is near impossible to catch this stuff "live" log the output to a file then use tools like grep to test for problems ... that's what I do ...
User avatar
By rudy
#57414 Sure. That is better. But I didn't plan to get too deep into the code. My head would explode.

I had first had experience with Coopdis's easyWebServer. He claimed better performance because of his native rather than Arduino calls. I spent some time looking at it and found some problems. But he didn't seem interested in fixing anything. I guess he had something new to work on and wasn't interested in cleaning up issues. I had been interested in easyMesh for a lot of reasons but I could not get it to compile without errors. Same problem a number of people had. I finally got back to it and got it to compile on my computer. It had problems but if those could be fixed it looked like it had potential. Especially with a single SSID approach recommended. But nothing to be seen of Bill/Coopdis. No explanation of guidance. No comment whatsoever.

But Scotty took the bull by the horns to stabilize it. Discover the weaknesses. Change what needed changing. And it is working better but still not there yet. I don't know how far he will take it or if he will take what he learned and start fresh. There still is a need for a working ESPmesh.

I don't have the time to learn the ins and outs of the code. Because I don't have the programming background and skills. I do hope to improve in time but at this time I am better at trying to break things (testing) rather than fix things.
User avatar
By picstart
#57435 Perhaps what we need with the Mesh is a defined Mathematical strategy.
An example could be

The solution space is asynchronous.
No device has a timing preference above any other.
( The original EasyMesh code imposes a master timer synchonization.)

Every device is equal
( All devices need to know the mesh structure and the list of devices active in the mesh.)
( messages have the freedom to be sent to any specific device and broadcast to all devices)
Each device needs to know its spacial position in the Mesh.
( Every device knows the radio strength (RSSI) it has with its neighbors)

Half duplex
esp8266 radio component can't receive its own signal.

Principal of Least Action
A message is delivered with the least radio action.
( if the RSSI permits a direct path between device the message goes direct otherwise the most efficient relay path is used)
( if a device isn't within radio range a broadcast is sent to devices in range to see if they might have that device within their range
if so an intermediate relay path is constructed)

With respect to a message a device is either an originator , recipient or a re-layer of a message

Confidence
With respect to a message the originator needs confidence the message was received by the recipient.

Compatability
Protocols UDP or TCP are required to retain Wifi compatability.