I've been getting about 55bps - that's fifty five bytes per second.
Is that to be expected? For some reason I assumed it would be at least 10x that. I'm using the Adafruit esp32spi libraries in CircuitPython until we get MicroPython support but would like a sanity/expectation check.
What was the speed other users were managed to get through an esp32?
Are there other WiFi systems in your area? They could kill the bandwidth. Also try it with a laptop and run the speed test. Is the system you are measuring your 55 bps capable of more? Test it with another system.
I did a quick test with the Arduino Nano RP2040 Connect connected to my WiFi router as a Webserver sending a web page to my Windows laptop connected to the same network. I did use a C++ example. So, it is not a fair comparison but maybe gives you a reference point.
The page is a single HTML around 400kB and took 17.8 seconds to load. During that time Windows network statistic showed 200kbit/s received and 8kbit/s send.
Wireshark shows that multiple prints where combined into IP packets (data size 1000 to 1400 bytes) which I guess is good for throughput.
And all of them return with 55 bytes per second using wget at the command line. I tried different sizes thinking maybe the python libraries were only sending chunks per write (with an overhead per write or something).
Thanks again. At least I now know it goes faster than 55bps
As a kind of funny update, I hacked together a CircuitPython firmware with the clockspeed doubled (250MHz). The download rates scaled (55bps at 125MHz, 130bps at 250MHz). So there we are - it's the "python tax".