Redundant info printed everytime a script is started

[jetpax]

Each time a script is run using the RUN button redundant information is printed before and after

eg

MicroPython v1.26.1-dirty on 2025-10-07; Generic ESP32S3 module with ESP32S3
Type "help()" for more information.
>>> 
raw REPL; CTRL-B to exit
>OKHello, MicroPython!
>
MicroPython v1.26.1-dirty on 2025-10-07; Generic ESP32S3 module with ESP32S3
Type "help()" for more information.
>>> 

Expected would be only to print banner on start up, and on each script run only

Hello, MicroPython!

cf ESP32 MPY-Jama

[ubidefeo]

@jetpax
currently all the stream over the serial port is handled by a buffer that is shared with the x-term instance.
I have a branch I've been working on to filter what goes in and out, but it's not stable/reliable yet.

Definitely adding this as a feature request, hope I can spend some time on it once I finish other more urgent tasks.

Thank you for pointing this out :)

[Josverl]

For the ESP boards I see this as an indication that the board has restarted for some reason, likely a spike/change on the DTR/RTS lines

[ubidefeo]

@Josverl
The board is not reset on Run, I use ESP32 a lot, can you elaborate on your thoughts? :)

[Josverl]

Look in the MicroPython repo for the fixes in Mpremote serial by me and @projectgus to avoid DTR/RTS spiking on Windows, and thereby inadvertently resetting MCUs , or dropping then unintentionally into bootloader.

To see it in action hook up a logic analyser to the UART and DTR RST pins.
And then do what you do normally.

As an example, before my initial fixes stopping more more caused all ESPxxx boards to reset.
Due to timing That only showed up in scripting, not in manual test, and the assumptions were human error, or bad drivers.
Reality was that the root cause was that Windows serial ports follow the same logic and conventions as linux, so essentially an in compatibility that pyserial does not solve or bridge.

This is not limited to Python though, Ive worked on similar patches for Java and Electron native serial ports.

[projectgus]

FWIW, it could be DTR/RTS triggering hard reset but then I'd expect to see some of the ROM bootloader log in the output as well. Are you opening and closing the serial port when this happens?

My first thought is this looks like a MicroPython soft reset is being triggered in interactive REPL mode (which prints the banner at the end the soft reset). If you trigger a soft reset in raw REPL mode instead then you won't get the banner (although it still prints something.) Play around with Ctrl-A, Ctrl-B, Ctrl-D in a mpremote or serial terminal to see what I mean.

(Quick look at micropython.js shows that the run method exits raw REPL mode at the end, and the reset method doesn't go into raw REPL mode, which matches my initial hunch although I'm not looking at all closely.)