Is this project still under continuous development? Is there any information about future design plans？

[BenDerPan]

This project is very interesting, if it can be integrated with the characteristics of current AI frameworks, such as Agent, MCP, etc. It seems interesting to have native support and integrate traditional functions with AI capabilities. Although there are still many difficulties in promoting a new language.

[emil14]

@BenDerPan

Hello and thank you for your interest!

Is this project still under continuous development?

It's hard to say. Let me explain...

07.04.2025 development was officially "frozen" by me as a main maintainer, there was a message on discord server.

However, my position was (and still is) that I encourage any contributions and ready to provide help to anyone, who want to participate. It's just that I myself don't have enough time to continue development.

BUT few days ago I actually decided to continue my work on tagged unions pull request, a massive language change that affects almost everything.

Implementing tagged unions is an important milestone that blocks a lot of other critical stuff such as pattern matching, which is super important for stdlib components and other language features.

So, what is the actual status? I would say it's like this:

I'll try to at least finish tagged unions PR because I just don't feel good because it's current state. Too much changes was made and not merged with the main branch. It's not right. However, I can't guarantee any schedule now. I'm still open to review contributions and help others thought.

Last thing I must say, is that no matter how busy I can get these days, eventually work on Nevalang will be continued, either by me or other people.

Hope it helps.

Is there any information about future design plans？

Yes, I would say I have clear vision for this project. You can learn about it here.

It's somewhat outdated because I stopped working on Nevalang for several months but it still can give you a clue of what's going on.

if it can be integrated with the characteristics of current AI frameworks, such as Agent, MCP, etc. It seems interesting to have native support and integrate traditional functions with AI capabilities. Although there are still many difficulties in promoting a new language.

I'm not sure if I fully understand what's on your mind. I can say that I am interested an applied AI. I actually have small Telegram channel about vibe-coding practices and the company I work for at the moment is doing GenAI integrations mostly. So I'm into this stuff.

However, in my opinion Nevalang is a programming language. I'm pretty sure at this point that, for example, syntax-level must not have anything AI related. Having some stuff in stdlib for example, to make it easier to build things, that's an interesting take.

Feel free to unwrap your thinking, I would love to discuss!

[BenDerPan]

@emil14 Thank you very much for your reply. Ultimately, Neva aims to become a programming language that supports business logic orchestration. I've reviewed the roadmap, and there are plans to implement it as a statically compiled language capable of running on any machine. In my opinion, the core of an orchestration language lies in supporting the procedural orchestration of various business application scenarios. In many cases, such orchestration needs to be dynamic to better adapt to diverse situations. This sounds somewhat similar to a dynamic workflow engine. If dynamism is required, wouldn't it be more convenient to design it using a virtual machine or a scripting/interpreted language? Or perhaps Neva could support dynamic parsing and execution of scripts in the future? If it's static, the advantages don't seem particularly obvious to me—it feels similar to visual programming frameworks like Scratch or Snap. I'm not sure if my understanding is correct. 😊

[emil14]

@BenDerPan

Neva is statically typed programming language, like Go or Rust, but in different paradigm.

The goals are:

1. Create a tool with better developer experience, make programming easier, more fun
2. Explore possibilities in programming language design

Now let me expand on this

Create a tool with better developer experience

Things like static analysis and visual programming goes here. Static analysis specifically should eliminate wide range of possible mistakes and free you from thinking about those things. And concurrency by the way is hard, so it's good to have compiler that can watch your back :)

Explore possibilities in programming language design

The controlflow paradigm is somewhat "solved" in programming languages. We have very good picture of what it is, what it's capable of, etc. On the other hand there not many dataflow languages and, what's important, not a single one like Nevalang. Statically typed, implicitly concurrent/parallel, with possible support of visual editor, etc. From what I see, many and many problems of "traditional" languages might be solved on a different level.

Last, but not least. Nevalang, as I see its future, should allow to solve specific class of problems easier that we do with controlflow languages. Everything that has something to do with concurrency. Controlflow just not go well with async stuff.

Now I woild like to ask you a few questions.

Neva aims to become a programming language that supports business logic orchestration

In the same way Go or e.g. Python supports it. Programming language is quite low-level tool so it's flexible to support wide range of scenarios. But I feel like I don't quite understand what exactly do you mean. Perhaps I don't have enough experience with some tools, this is my intuition now.

In my opinion, the core of an orchestration language lies in supporting the procedural orchestration of various business application scenarios. In many cases, such orchestration needs to be dynamic to better adapt to diverse situations. This sounds somewhat similar to a dynamic workflow engine. If dynamism is required, wouldn't it be more convenient to design it using a virtual machine or a scripting/interpreted language? Or perhaps Neva could support dynamic parsing and execution of scripts in the future?

Depends on what do you mean by "orchestration language". It's a programming language, first of all. Also depends on what kind of dynamism is required.

There are some ideas about reflection and dynamic changes to the program's graph, I believe, that are possible to find somewhere inside this project, somewhere in issues and discussions but that's far from first priority for now.

Hope it helps, and thank you for your response!