DEV Community

Muhammed Shafin P
Muhammed Shafin P

Posted on

On NDM-TCP, Open Source, and Ethical Concerns

A statement on licensing, commercialization, and responsible use


What This Article Is About

This document clarifies:

  • What NDM-TCP is intended for
  • My position on how it should (and shouldn't) be used
  • Concerns about potential misuse

Important caveat: Many of the concerns raised here are only applicable IF NDM-TCP performs well in real-world scenarios and on real hardware, beyond the current localhost simulation tests. If it doesn't translate well to production environments, these concerns may be largely theoretical.


My Position on Copyright and Ownership

Since NDM-TCP is open-source and my hobby project, I don't care about copyright or related ownership issues in the traditional sense. I don't want control over it.

The code is released under GPL v2—not because I have strong feelings about that specific license, but simply for compatibility with other projects. That's all.


The Real Concern: Hardware and Implementation Loopholes

Here's what keeps me up at night.

If NDM-TCP really proves that it works on real hardware, the implications go far beyond the GPL-licensed code:

What GPL Doesn't Protect Against

A company could:

  • Build custom hardware with the soul (idea or whole concept) of NDM-TCP embedded in it
  • Create their own version from scratch in C, Rust, eBPF, or any other language
  • Implement it in firmware or specialized networking chips
  • Make it proprietary and closed-source

GPL doesn't protect against this. You can't copyright an idea or a concept.

The Potential We Can't Even Imagine

This technology has potential in areas I haven't even thought of yet:

Military applications:

  • Tactical network optimization in contested environments
  • Priority traffic management in defense systems
  • Resilient communications under attack conditions

Commercial weaponization:

  • Unfair competitive advantages in CDN/cloud infrastructure
  • Proprietary networking ASICs with NDM-TCP principles baked in
  • "Secret sauce" that gives one provider an edge over others

It's a double-edged sword.

Where I Draw the Line

If you're using this for:

  • Open-source projects
  • Public study and research
  • Academic purposes
  • Community benefit
  • Making the internet better for everyone

It's free to use as you want. You have my full support.

If you're using this for:

  • Military applications
  • Proprietary commercial advantages
  • Closed-source hardware implementations
  • Competitive weapons against other companies
  • Purposes that harm fairness and openness

I don't support it, and I hope the community doesn't either.

But I acknowledge I can't stop it. That's the nature of publishing ideas.


You Can Build On It—But Consider the Ethics

About the Code Quality

I need to be transparent: I use AI to code—specifically Claude Sonnet 4.5 and Gemini. This means the code isn't that much optimized.

If the community needs an optimized version, you can optimize it further. If there are advanced developers out there who see the potential and want to make it production-ready, please do. The code is there as a proof of concept, not as the final word.

Freedom to Build

You are absolutely free to:

  • Build upon the ideas I've used
  • Create your own implementation based on these concepts
  • Fork the code and take it in new directions
  • Use it in your projects

That's what open source means, and I stand by that.

But I ask you to think carefully about how you use it and what you use it for.


The Double-Edged Sword of Innovation

Network congestion control is inherently dual-use technology. It has potential we can't even think of yet.

Positive applications:

  • Improving internet performance for everyone
  • Making video calls smoother in poor conditions
  • Helping remote workers in underserved areas
  • Reducing latency for real-time applications
  • Making networks more efficient and fair

Concerning applications:

  • Giving unfair advantages in competitive commercial products
  • Creating proprietary "secret sauce" that locks users into platforms
  • Potentially enabling traffic manipulation in ways I haven't foreseen
  • Use in specialized networking contexts with ethical implications

Why I'm Publishing Anyway

Despite these concerns, I'm making NDM-TCP open and available because:

  1. Transparency serves the community better than secrecy. If there are problems with the approach, the community can identify and address them.

  2. Good ideas should be freely available. Locking this away helps no one.

  3. Trust in collective wisdom. I believe the networking community, as a whole, will use this knowledge responsibly.

  4. Academic and hobbyist advancement. People studying TCP congestion control deserve access to experimental approaches.


A Request to the Community

If you use NDM-TCP or build upon these ideas:

Please consider:

  • Will this make the internet better for everyone, or just better for your customers at others' expense?
  • Are you creating proprietary advantages that fragment the open internet?
  • What are the downstream effects of your implementation?

Please avoid:

  • Using this as a weapon against competitors in ways that harm end users
  • Creating closed-source derivatives that hide what you're doing to traffic
  • Applications that could harm people or enable unethical behavior

My Hope for This Project

I created NDM-TCP as a crazy experiment to solve real problems that networks face. My hope is that it:

  • Contributes to the body of knowledge around congestion control
  • Inspires better solutions (whether based on this code or completely different approaches)
  • Gets tested, critiqued, improved, and validated by the community
  • Makes networking better for everyone, not just a privileged few

If it turns out that NDM-TCP performs well in real-world scenarios, I want that performance to benefit the open internet—not to become a proprietary secret or a competitive weapon.


Final Thoughts

I'm a student working on a hobby project between exams. I don't have a company. I don't have patents. I don't have commercial interests.

What I have is an idea that might work, and a belief that ideas should be shared.

Use it wisely. Build on it freely. But please, think about the impact.

The internet works best when innovations benefit everyone, not when they become tools for competitive advantage at the expense of fairness and openness.


NDM-TCP is available under GPL v2. Use it, modify it, learn from it—but consider the consequences of how you deploy it.

Top comments (0)