Merging qumat-qdp into qumat

[guan404ming]

Current Situation

Related #875

• qumat: Pure Python package (hatchling build)
• qumat-qdp: Rust extension package (maturin build)
• Installation: Users install via pip install qumat[qdp] for full functionality.
• Maintenance: Two separate packages, two separate release processes.

---

Option A: Merge into Single Package

Pros	Cons
Simple install: just pip install qumat.	All users must download Rust extension (even if unused).
Version Sync: Always in sync; no compatibility issues.	Must build wheels for every platform (Linux/macOS/Windows × x86/arm).
Unified Workflow: Single repo, single release process.	CI/CD becomes more complex (maturin + multi-platform builds).
Better UX: More intuitive import paths.	Pure Python users forced to wait for compile or download large files.
	Unsupported platforms cannot install (no fallback).

---

Option B: Keep Two Packages (Current)

Pros	Cons
Lightweight: Pure Python users get a fast, small install.	Discovery: Users need to know to use pip install qumat[qdp].
Flexibility: Can release independently with separate versions.	Overhead: Must manage version compatibility manually.
Robustness: Graceful fallback when a platform is unsupported.	Maintenance: Two distinct packages to maintain.
Efficiency: Small wheel size for the Python-only part.
Modular CI: CI/CD pipelines can be separated.

---

[rawkintrevo]

I was the one who originally raised the flags in #875 but now that I see pros and cons cleanly spelled out, I see advantages to current.

[andrewmusselman]

I agree, let's keep it how it is.

[guan404ming]

I agree as well!

[ryankert01]

Agree!

[rich7420]

Agree~

[guan404ming]

I will open pr to document detailed reason in ADR

[rawkintrevo]

Afaic this github discussion can serve as the adr, i've seen some project use the wiki for ADRs too