IMPROVDEBUG:_168_(invalid) in log

[flashmark]

I am working on a plugin which was running stable for several days. Few days ago I updated my branch to latest mega version. Since this update the log shows many entries with "IMPROVDEBUG:_168_(invalid)". No changes in my own code were made. and the device is sometimes unresponsive. I am running a WT32 -ETH01 with custom build.Any suggestions where to look for a root cause of this behavior?

I do see commits for ImprovWiFi of last week. However I am using the Ethernet connection of the device. Maybe this use case was not foreseen/tested with these changes?

[Ath]

What does your plugin do to trigger the IMPROV serial handler? Got some 'handle' method that's not bound to a specific class?

[flashmark]

I deactivated all plugins (including my own work in progress) and deactivated all controllers (I had Domoticz HTTP in use).
There is no intended connection at all between the plugin and any serial handler. The plugin reads a value from another plugin and switches an GPIO digital output based upon state and input value. The plugin does use the PLUGIN_WRITE to read a remote on/off command.
So I assume it is either triggered by the fact I am using the Ethernet connection instead of WiFi or a unintended correlation like memory issues. But by disabling the plugin I wonder if it can trigger such a situation. So I am loking for suggestion how to debug this, for me unknown IMPROV serial handler.

[Ath]

The message you have shown is only found in the IMPROV handler, that's why I asked.

Most likely you have some unclosed structure, or a missing break somewhere in your code, or are calling some uninitialized class method, that accidentally points somewhere in the IMPROV code .

The plugin reads a value from another plugin and switches an GPIO digital output based upon state and input value. The plugin does use the PLUGIN_WRITE to read a remote on/off command.

Hm, don't yet understand what you are trying to achieve there...

So I assume it is either triggered by the fact I am using the Ethernet connection instead of WiFi or a unintended correlation like memory issues. But by disabling the plugin I wonder if it can trigger such a situation. So I am loking for suggestion how to debug this, for me unknown IMPROV serial handler.

That's IMHO unlikely, as IMPROV has nothing to do with the Ethernet configuration, but is intended as a means to configure the device via the serial port.

[TD-er]

Without posting the code it is near impossible to give a good suggestion where this might go wrong.
Like Ton mentioned, IMPROV is used to interact via the serial port when you have the ESP connected to your PC and try to connect with it via the web flasher.
The IMPROV code does only interact with the serial port if a message has been received starting with IMPROV, which makes it highly unlikely you will every trigger it by accident.

[flashmark]

Sorry I could not work on this for some time. Today I discovered the issue. Somehow the combination of WT32-ETH01 and enabled serial port does not work and triggers this logging. It also makes the module quite slow on the web gui. Strange it seems to occur after a rebase to a newer baseline. Thanks for comments and explaining the message. Thsi discussion can be closed.

[Ath]

Are you working on a new plugin with ID P168? As I'm about to assign a number to another planned plugin, but if you are already using that I can also register yours and select the next ID to my planned plugin. Just provide a name and purpose for your plugin.

[flashmark]

Hi Ath, no currently the plugin is P254 as place holder, no intention to deliver it on short notice. I am still in doubt to combine it with an existing plugin P021 altough my original setup is quite different. I will post a message when I need help to decide. And I will of cause first reserve a plugin ID using the github ticket used to that.