Matthew Kennedy

``` rusefi/unit_tests$ rm -rf .dep build rusefi/unit_tests$ time make -j20 .... real 1m52.901s user 6m27.688s sys 7m25.625s ``` then when touching one file (the "dev loop"): ``` rusefi/unit_tests$ time make...

This is using WSL (I think it's even WSL v1), btw. It's significantly faster still on my slower-hardware-but-actual-Linux machine.

real linux machine says clean build: 1m5s touch main.cpp: 6.7s

setting `USE_LINK_GC = no` on WSL machine says clean: 1m39s main.cpp: 12.3s

Is that on master? 7 seconds seems pretty good to me.

so problem solved?

1. so perhaps a better issue would be "tear down `Engine` god object" 2. you can `wsl bash -l path_to_script.sh` from anywhere in windows and it'll do the right thing,...

Just did another test, looks about the same as last time for the dev-loop case. ``` $ make clean $ time make -j12 ..... Linking build/rusefi_test real 1m25.655s user 7m42.292s...

> linking performance is not rusEFI problem but NTFS problem? this is a well known problem - NTFS is good at some things, but running a compiler (or linker) is...

There do seem to be two ways to do "use only rising": - the trigger can set it (`useRiseEdge`) - the user can set it (`useOnlyRisingEdgeForTrigger`) and they seem to...