Srinivasa Mahesh

The connection doesnt work(all reads from the pipe used to communicate return 0 bytes) if chrome is already running with the same user_data_dir. I need to investigate further.

Your proposal seems interesting. I will think about it when I get more time. Regarding interfaces, they can only verified at runtime so is that OK?

I think the best way will be to have two types of equality. In Javascript Object.is(-0,0)==false and Object.is(NaN,NaN) == true but -0==0 and NaN!=NaN

A similar situation exists for NaN. https://wren.io/try/?code=G4QwTgBADgdiMF4ByBXAtgOjjAUKSM2CAtKptjgMoCeAzgC4CmmUYAljPQBSzwCUVOkxbtOXQv0ENmGVh24B5AEYArRgGN6GWiDSMe2ADS8YfATWkj5XZWs3bd+k4YmmBQA `Object.same(_,_)` treats positive and negative nans differently.

I managed to use pipes in my library https://github.com/Srinivasa314/alcro in the master branch, but you need to use cgo for it. ~~Also headless mode does not work (I dont know...

UPDATE: I fixed a bug in version 0.1.4 so I yanked it. Also I changed the version of winapi from ^0 to ^0.3. Documentation for 0.1.5 appears only for linux...

There are many bugs in this pull request so I created #32. It also fixes the problem mentioned by @jakiki6 .

> 1. input delay / strange input behaviour I am not entirely sure. This is what I think. The player continues to move for a while after input is stopped...

> I found that compiling on mac I had to change the -lGL flag to -framework OpenGl Fixed it > Even then though it compiled fine but I got segmentation...

You can use `lldb` or any other debugger to get the stack trace. Add -g to CXXFLAGS in the Makefile before compiling so that more information will be available to...