Tim Whiting

The issue is that you are not actually showing what is in the ref, and never using the parameter: It infers: ```koka pub fun ref/show( r : ref, ?show :...

Thanks, this is something we can fix easily.

This is something that we've run into on the C backend as well (name clashing - not the stack overflow related to this). We need to likely come up with...

It looks like we are setting both the locale and code page to UTF8 on process init. https://github.com/koka-lang/koka/blob/eb83468fdf9f0f6a7d1f22471f92844c9f0dc9f0/kklib/src/init.c#L179 https://learn.microsoft.com/en-us/windows/win32/intl/code-page-identifiers https://learn.microsoft.com/en-us/cpp/c-runtime-library/reference/setlocale-wsetlocale?view=msvc-170 We might need to change this: ```C setlocale(LC_ALL, "C.utf8"); ```...

> but the advantage of tab literals is that you can visually check certain alignment properties of your output in the source code, especially when outputing tabular data. I think...

This is related to the ordering of effect rows. I was unaware until recently that you could scope named effects without using a separate effect for delimiting the scope. In...

Likely related to the type inference problems in #769 like you mentioned.

> Some questions: > Are effect rows "ordered"? (Or better question: should they be ordered from a programmer's point of view?) Apparently they are, but documentation doesn't really reflect that...

If you look at the linked branch in the PR description you should see how to use this with vectors. I'm not sure what other code you have going on....

Wow that is really strange. Yes, I wouldn't expect the array to be copied using unique-copy (that is the point of this PR). However, I see the same strange behavior...