mame82

> up to a 50:50 split That's why I stated one ends up with half-duplex (if both side are sending, although your idea scales very well, if only one side...

News: Replacing `AsyncWrite` on my Background `MemoryStreams` with `BeginWrite/EndWrite` dropped my Transfer rate into the expected range: `End receiving /tmp/test received 1.048.517 Byte in 41,2182 seconds (24,84 KB/s)` Every call...

> I struggle to believe that BeginWrite/EndWrite on a MemoryStream is necessary, or even efficient. Unless you are doing it to be compatible with Async operations on other types of...

Next interesting Observation: Starting only the powershell side of communication (no python endpoint on Pi Zero) it turns out that `write()` is blocking to. If no reports are read on...

@RoganDawes here're the test files https://github.com/mame82/tests/tree/master/fullduplex I guess I have to reimplement everything, as alternating read/write is the worst approach. Have you found a replacement for named pipes to Interface...

> The main issue then is failure to read packets on the Powershell side, resulting in lost data. This is easily seen by introducing a console write in the read...

Okay, as already discussed on Twitter, this approach was **invalid**. Patches for SDK code (namely `HttpClient`) do not end up in `libflutter.so` as the VM snapshot is compiled into `libapp.so`....

For completeness, I attach a compiled libflutter.so: - Snapshot hash: `9cf77f4405212c45daf608e1cd646852` - Platform: Android - Arch: arm32 - Patches: All reFlutter patches, `connect.cc` patches have Address rewrite disabled, but still...

T.b.h. even if I could remember what I did 3 years back, I would not give such a clear advise. Suggesting (how I would approach it, if I have to...

> This is still happening on all of our Windows machines with v5.4.0. Same after updating from 5.3.1 to 5.4.0 (win 11) ``` during RTL pass: ira C:/Users/xxxx/esp/v5.4/esp-idf/components/esp_lcd/rgb/esp_lcd_panel_rgb.c: In function...