@chrispsn I guess that if we stick to the lists are maps theory, cutting a dict should be no different to cutting a list, you would cut the domain and the range
I think that for consistency they should be integer indices, but I am not sure that would be useful for anything. k usually values more convenience than consistency
@ngn i'm trying to compile ngn with tcc. it complains that this: ci[i][2]=al(t[i].l) is a division by zero in constant. Any idea? maybe it's something completely different.
@yiyus we were dicussing a typo in you document: https://chat.stackexchange.com/transcript/message/53613885 is that the latest version (linked somewhere in hn)?
@yiyus thank's but i don't feel qualified. i only skimmed through it (via your doc) which looks like a major undertaking. due to kernighan, i should be twice as smart as arthur in order to understand it anyway.
@chrispsn maybe for symmetric differentiation ;)
@chrispsn it must be an important concept. some even call their company after it.
@ktye can't keep them all happy. use clang-7 or gcc 6.3.0. there's something like a c standard (many of them, actually) but i never learned it and it probably won't work well with this style of coding