bfredl

> w_cursor.col is a byte offset, not a column, despite being a colnr_T. Most other uses of colnr_T are actual screen columns. In general "column" actually refers to byte counts,...

> API for, say, splitting a window, but opening a different buffer in the split, where the second buffer's options are used for the new window instead of the window...

> It's maybe off-topic here though, is there some open bug or anywhere more appropriate to discuss this? Not what I know of. Most recently we have discussed this on...

doesn't seem to fully fix it in my original repo. I suppose we need to resolve `@sizeOf(S) > unsigned_runtime_val` as well? (but not with `==` ? or is just comptime...

I guess the root of the question is under what circumstances the following kind of code, which exists in the test suite, is okay. ``` test "@sizeOf(T) == 0 doesn't...

``` test "foobar" { const S = struct { const Foo = struct { y: if (@sizeOf(Foo) == 0) u64 else void, }; }; try expect(@sizeOf(S.Foo) == 0); } ```...

> Why is this specific to signed integers and lhs being a lazy value? Swapping the order or making the other type unsigned immediately breaks again. @Vexu it was suggested...

I posted two examples above. https://github.com/ziglang/zig/pull/12539#issuecomment-1221423767 is currently in the test suite and looks fine in isolation. But also https://github.com/ziglang/zig/pull/12539#issuecomment-1221425691 passes and I don't think it should. I don't really...

> but is it still the case that neovim's internal screen is composed of unicode characters of fixed size and hence will never be able to display images. "never" is...

> A single "sixel" encodes information about a six pixel high * one pixel wide column. So each sixel in binary is encoded as a single 6-bit number. This is...