THE BRIDGE BETWEEN HARDWARE WALLETS AND BITCOIN CORE IS BEING RETIRED — NOT BECAUSE IT BROKE, BECAUSE PYTHON CANNOT BE BUILT THE SAME WAY TWICE — NO NEW DEVICES ACCEPTED AS OF NOW
If you plug a hardware wallet into Bitcoin Core, or into Sparrow, or into most of the desktop software people use to hold their own coins, the thing that makes that conversation possible is called HWI. On August 18 its maintainer, Ava Chow, opened an issue titled “Future of this repo” and said it is being wound down.
Read the reason, because it is not the one you would guess. Not money. Not a fight. “Limitations of Python prevent us from going all the way as we are unable to deterministic builds or ship it with Bitcoin Core.”
Bitcoin Core ships in a way anyone can check. You compile the source yourself, and if what you get does not match the released file byte for byte, something is wrong and you can prove it. That is the entire point. Python cannot do that reliably. So the tool that connects your hardware wallet to Core could never actually ship inside Core, and after years stuck at that wall, the project is stopping.
Here is the part that reaches your desk. Effective now, no new device support is being accepted. Finish one feature, one last release, then archive. Buy a hardware wallet built from here on and nothing promises the old bridge will ever learn to speak to it.
A replacement is starting over in a different language. It is marked work in progress.
Nothing was hacked. Nothing failed. A useful thing is being retired because it could not be proven, and proof is the standard. That is not a flaw in Bitcoin. That is Bitcoin working, and it still costs somebody a tool. Technologist.
Read the reason, because it is not the one you would guess. Not money. Not a fight. “Limitations of Python prevent us from going all the way as we are unable to deterministic builds or ship it with Bitcoin Core.”
Bitcoin Core ships in a way anyone can check. You compile the source yourself, and if what you get does not match the released file byte for byte, something is wrong and you can prove it. That is the entire point. Python cannot do that reliably. So the tool that connects your hardware wallet to Core could never actually ship inside Core, and after years stuck at that wall, the project is stopping.
Here is the part that reaches your desk. Effective now, no new device support is being accepted. Finish one feature, one last release, then archive. Buy a hardware wallet built from here on and nothing promises the old bridge will ever learn to speak to it.
A replacement is starting over in a different language. It is marked work in progress.
Nothing was hacked. Nothing failed. A useful thing is being retired because it could not be proven, and proof is the standard. That is not a flaw in Bitcoin. That is Bitcoin working, and it still costs somebody a tool. Technologist.
BITCOIN-CORE/HWI ISSUE 850 (PRIMARY) · opened by achow101, Tue Aug 18 — two weeks old, picked up since · + BITCOIN OPTECH #420 Fri Aug 28 · + CRYPTOSLATE on the wallet-side consequence · existing devices keep working; only new device support is closed