thank you for the feedback!
I have found out what is wrong with compilation for ESP32 Dev Module.
The ESP32 processor hangs on a second call to Serial.begin(115200) when compiled for ESP32 Dev Module.
// SerialBeginTwice test by Jean-Marc Zingg
//
// Purpose: test issue with ESP32 compiled for ESP32 Dev Module
void setup()
{
Serial.begin(115200);
delay(100);
Serial.println("going to call Serial.begin(115200); a second time. Should tell success ...");
delay(100);
Serial.begin(115200);
delay(100);
Serial.println("... success, returned from second call to Serial.begin(115200); Bye and out.");
delay(100);
}
void loop() {}
I verified this on a ESP32 Dev Module clone, on a Wemos LOLIN32 Lite, and on the Waveshare e-Paper ESP32 Driver Board.
The GxEPD2 examples call Serial.begin(115200) in setup() to Serial.print("setup"), and again in display.init(115200).
The example GxEPD2_WS_ESP32_Driver.ino with GDEM029T94 works ok if compiled for WEMOS LOLIN32.
(I first had an issue with garbage during partial update, because I had selected GxEPD2_3C instead of GxEPD2_BW).
The display class and the driver class need to correspond for monochrome or 3-color.
I don't see an easy way to check during compile time. Maybe I should add a warning diagnostic at runtime.
Jean-Marc