Thank you for your interest in D-MemFS. To maintain the "bulletproof" nature, "zero-dependency" design, and architectural purity of this project, we have established a stability-first contribution process.
Before proposing any changes, please read the Compatibility and Non-Goals section in the README. Understanding the intentional limitations of D-MemFS is a prerequisite for any contribution.
To minimize review overhead and ensure alignment with our "Spec-First" development model, we follow a strict Issue-First workflow:
- Open an Issue or Discussion first. All bug reports, feature ideas, or documentation improvements must start as an Issue or a GitHub Discussion.
- Wait for consensus. We will discuss the proposal and determine if it aligns with the project's design principles and current roadmap.
- PRs by invitation. Pull Requests should only be submitted after a consensus has been reached in the corresponding Issue/Discussion. Unsolicited PRs, especially those that bypass the design discussion, may be closed without review to maintain focus.
- Bug reports — please include a minimal reproducible example, Python version, and OS.
- Documentation — improvements to clarity, fixing typos, or adding useful examples.
- Security concerns — direct reporting of potential vulnerabilities.
For ideas that push the boundaries or involve features currently listed as Non-Goals, please use this dedicated space: 👉 Discussion #2: Boundary-pushing ideas and Non-Goals
Any accepted proposal must adhere to these three pillars:
- Zero-dependency — relies only on the Python standard library.
- Hard Quota consistency — must not bypass or weaken memory enforcement.
- Portability — platform-agnostic, requiring no administrative privileges.
git clone https://github.com/<your-username>/D-MemFS.git
cd D-MemFS
uv pip compile requirements.in -o requirements.txt
uvx --with-requirements requirements.txt --with-editable . pytest tests/ -v --timeout=30D-MemFS is currently in a stability-first phase. We prioritize architectural purity over rapid feature expansion.