You can tell lowsync to exclude the datas directory from syncing, which you should do, because otherwise you have it twice on the microcontroller.

Let me look up the setting in the next days...

I apologize for the problem.

Please try out these instructions: Fixing Updater Issue

Let me know whether this works for you.

Thomas

Whenever it has Internet access. But, as you opened the other topic regarding this, I guess you already figured that out

Regarding (1): Yes, the IDE file system view is not updated automatically. This is on the feature request list, but not much on the top.

Regarding (2): The Start Button thing is more urgent.. To be honest, I haven't tested the IDE much after implementing static files.

Both are put on the TODO List!

code.console_kb = 0 means it does not save the console messages for a while so you cannot see messages when you reload the IDE for example. But this is much faster as it does not have to write files. Much better for production.

Actually, I can' tell directly as I do not remember myself
I will try to reproduce this problem this weekend and keep you posted.

Correct. Also see https://github.com/neonious/lowjs_esp32_examples/blob/master/drivers/sensors/nmea_gps.js

Interesting. Can you try to set lowsync to use http, not https?

factory_files is only copied on --init, not on future flashes without --init, here you will want to keep your data files.

I really need to replace the SPIFFS file system. If it does not find a file, it searches through every sector. Not really feasable for Node.JS which looks in many locations. I have on the list to switch to littlefs, that should help with this problem. Till then, good to hear we have a solution.

It works well here. The not very often happening checksum problem I do see as something which could be coming from the microcontroller. Maybe exactly during a pulse an interrupt broke the measurement. But the not connected problems are weird, there is no pulse at all.

Not connected for long periods at a time sounds like a connection issue to me. Not sure why the microcontroller should not work for a few secs... A flacky cable connection could be it however. How have you connected the items? With a breadboard?

Alternativly it could be that the pulse to start the DHT11 is too long. I do not know if there is a maximum length for the pulse. Hopefully not, then it does not matter if JavaScript adds additional delay.

I believe more in the first reason. Can you check everything? Maybe rebuild the circuit on another board with other cables? If that does not work, I will have to try on a neonious one (tried on a different board, but that should not change anything).

Thanks