[mtinkerhess]

A lot of people are replying saying this isn't a big deal in practice. I believe them but I'll chime in and say this is also a dealbreaker for me.

I would love to find a language for game / engine development that's as performant as C++ and as easy to use as C# or Python, but this issue scares me. If I search for a variable using any of my existing plain text search tools, I expect to find the variable. It makes me think that this isn't a language that's meant to be used by large teams on large, long-lived codebases.

[dom96]

> If I search for a variable using any of my existing plain text search tools, I expect to find the variable.

I think you are over-estimating how inconsistent Nim code typically is. I have been coding in Nim for a very long time and I personally just use case insensitive search (which I use for all languages anyway).

It's easy to tell from context what the identifier will be. If the code base is snake_case then I search for `foo_bar` if not then I search for `fooBar`. I cannot remember a single time this has failed me, and even if it does then the worst case is you'll need to search for the other alternative. There is really no trouble here and I feel like a lot of people, when they see this feature, worry too much because it's so different. Please please please reconsider, I know it's a strange feature and it doesn't feel worth it, but it really isn't a problem at all.

[mtinkerhess]

I totally hear you that this is not a huge inconvenience in practice. Part of me is worried that it becomes more and more of a problem the bigger the project gets.

But mostly, it just seems like a really weird call to make for the language. There are real disadvantages to it -- I can't really rely on my search tools. So that... when you're using a library you can refer to its identifiers using a different capitalization convention if you want to--why is that important?

I don't know, it just seems like a really weird call to make, and if that's one of the first things I've learned about the language it makes me think there are probably plenty of other weird things like it.

[dom96]

> Part of me is worried that it becomes more and more of a problem the bigger the project gets.

But why? I personally haven't experienced this having worked in a big Nim project. I'm really curious why you think a bigger project could become burdened by this.

> There are real disadvantages to it -- I can't really rely on my search tools.

I disagree. You can rely on your search tools, as I've demonstrated in my previous reply. I feel like you're thinking of the worst case scenario and convincing yourself that it's very common. It simply isn't, people don't create crazy identifiers, why would they?

> So that... when you're using a library you can refer to its identifiers using a different capitalization convention if you want to--why is that important?

It's important for consistencies sake. In the real world developers prefer different conventions, so you undoubtedly run into a library which uses a convention that is different to your own. It's nice to be able to call that libraries functions in the same convention as your own.

I agree that it's not a huge advantage, which is why I created this thread on the Nim forum recently: https://forum.nim-lang.org/t/4388. My main reason to consider getting rid of it though is people like yourself who dismiss the whole language because of this small feature :(

[Recurecur]

I'd take a long, hard look at Julia if I were you.

It has a lot in common with Nim, but is backed by a much larger community. Its FFI interface for C functions is faster than...C.

If Julia doesn't fit your use case(s), Rust might also be worth a look. What worries me about Rust is that it looks to me to have an awful lot of boilerplate that detracts from readability.

[alehander42]

You aren't supposed to use different styles in the same codebase: Nim is going to have a tool that can enforce this soon.