make menuconfig -> component config -> ESP32-specific -> UART for console output

:lol: esp32 keeps us guessing. Do you have to Max out the drive strength setting to get there and what is the max current per domain? And what about rtc I/o mode? Yes, as the datasheet specifies, this is at max drive strength. RTC I/O is the same, I presume. We indeed also need to specify the curre...

To be somewhat nitpicky: you'd still use the RTOS as the socket layer and the WiFi stack are depending on it. But if you're asking on how to do this without creating tasks for each client: you'd use something like select() to block until there's activity on any of the clients, then handle them appro...

What hardware do you use, and did you make sure the hardware doesn't already use GPIO4 and/or 5 for the flash? Rerouting a flash pin would cause the software to crash like this.

Actually, no. There was some fuzz about the output current handling capacity for a while, but as far as I know, the 40mA number is correct: the 12mA number we used initially was a holdover from the ESP8266, but for the ESP32 we improved the IO-pads, so they can handle a larger current now.

You are more-or-less right. If I recall correctly, we can actually have some more psram (up to 8MiB or so) but it cannot all be mapped into memory at the same time, and as it'd become quite difficult to use, we have no plans to add functionality for this to esp-idf.

Locking: dup with viewtopic.php?f=13&t=5721

No, sorry, the PID system is more-or-less untested at the moment.

Hmm, 320*480*30fps*16bit is almost 80MBit/second... that's a fairly hefty data stream. I'm actually not even sure if the SPI bus is fast enough to handle this (unless you go for quad-spi) and I'm pretty sure you will not be able to maintain an 80MBit/sec stream in real WiFi circumstances.

I'm pretty sure you can use the RMT peripheral for that. The MCPWM block may also have a capture mode that allows you to capture pulse lengths, but I'm not 100% sure on that.