351 lines
19 KiB
HTML
351 lines
19 KiB
HTML
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
|
|
<HTML>
|
|
<HEAD>
|
|
<TITLE> [zapps-wg] Powers of Tau Ceremony Proposal
|
|
</TITLE>
|
|
<LINK REL="Index" HREF="/pipermail/zapps-wg/2017/index.html" >
|
|
<LINK REL="made" HREF="mailto:zapps-wg%40lists.zfnd.org?Subject=Re%3A%20%5Bzapps-wg%5D%20Powers%20of%20Tau%20Ceremony%20Proposal&In-Reply-To=%3C6CFF92D2-A041-492E-A2A1-69B4B00DA626%40jasondavies.com%3E">
|
|
<META NAME="robots" CONTENT="index,nofollow">
|
|
<style type="text/css">
|
|
pre {
|
|
white-space: pre-wrap; /* css-2.1, curent FF, Opera, Safari */
|
|
}
|
|
</style>
|
|
<META http-equiv="Content-Type" content="text/html; charset=us-ascii">
|
|
<LINK REL="Previous" HREF="000008.html">
|
|
<LINK REL="Next" HREF="000010.html">
|
|
</HEAD>
|
|
<BODY BGCOLOR="#ffffff">
|
|
<H1>[zapps-wg] Powers of Tau Ceremony Proposal</H1>
|
|
<B>Jason Davies</B>
|
|
<A HREF="mailto:zapps-wg%40lists.zfnd.org?Subject=Re%3A%20%5Bzapps-wg%5D%20Powers%20of%20Tau%20Ceremony%20Proposal&In-Reply-To=%3C6CFF92D2-A041-492E-A2A1-69B4B00DA626%40jasondavies.com%3E"
|
|
TITLE="[zapps-wg] Powers of Tau Ceremony Proposal">jason at jasondavies.com
|
|
</A><BR>
|
|
<I>Fri Nov 10 19:25:33 EST 2017</I>
|
|
<P><UL>
|
|
<LI>Previous message (by thread): <A HREF="000008.html">[zapps-wg] Powers of Tau Ceremony Proposal
|
|
</A></li>
|
|
<LI>Next message (by thread): <A HREF="000010.html">[zapps-wg] Powers of Tau Ceremony Proposal
|
|
</A></li>
|
|
<LI> <B>Messages sorted by:</B>
|
|
<a href="date.html#9">[ date ]</a>
|
|
<a href="thread.html#9">[ thread ]</a>
|
|
<a href="subject.html#9">[ subject ]</a>
|
|
<a href="author.html#9">[ author ]</a>
|
|
</LI>
|
|
</UL>
|
|
<HR>
|
|
<!--beginarticle-->
|
|
<PRE>Hi all,
|
|
|
|
Here is my report:
|
|
|
|
Powers of Tau Operational Writeup
|
|
=================================
|
|
|
|
Round: 2
|
|
Date: 2017-11-10
|
|
Name: Jason Davies
|
|
Location: London, UK
|
|
|
|
Challenge: 467bc84f6eb98ff956eaf12a1b7ef4dc0aff1093c7a0d5c1dfbdb85bbfffb20a43965d0daefee3fec6c1a47af69100e117b44b74371824ac8af1e33b6f91add5
|
|
Response: 2f728af894524f55bda7a3e2c2e2db6a57a992811e90ed57456d62aead5106cdc5c97c86532d14b5185cc74d169f1b0c2c0ef1e582231ffa7936da55047c0cb2
|
|
|
|
Preparation Steps
|
|
=================
|
|
|
|
Git repository: <A HREF="https://github.com/ebfull/powersoftau">https://github.com/ebfull/powersoftau</A>
|
|
Commit hash: 9e1553c437183540392a7231d0788318a19b18a3
|
|
Compiler: rustc 1.23.0-nightly (d6b06c63a 2017-11-09)
|
|
Build: cargo build --release --features=u128-support
|
|
b2sum(./target/release/compute): be42f68b07c5c857bb6561a9ac2967d671ef412a71c87c2fb31776a6ab38c756736de66e554553021e129ecab45d922092873df8b71bd9a775ec05f189485198
|
|
|
|
I used a brand new 16GB USB stick and loaded ubuntu-17.04-desktop-amd64.iso; b2sum: 6a1c975b25b4e7f2dbf4fda84fe8b5de3ed6f4532b8c4f17e533ed11a0a8b5b9ad9fb83e8e4b89447c3a427be73f77a5f7c71b7f733fcc4bebf346e9c5c0de43.
|
|
|
|
I reformatted a second brand new 16GB USB stick to ext4, then copied the
|
|
`challenge` file and the `target/release/compute` binary.
|
|
|
|
Sidechannel Defences
|
|
====================
|
|
|
|
First of all, I lined a large cardboard box with aluminium foil in order to
|
|
make a rudimentary faraday cage. Then, I assembled an airgap compute node
|
|
using some relatively cheap parts, putting them all inside the box:
|
|
|
|
* Motherboard: Asus H81 Pro BTC (no radio, bluetooth or speakers AFAIK)
|
|
* CPU: Intel G1840
|
|
* Ram: 2x cheap 1GB sticks
|
|
* PSU: EVGA SuperNOVA 1300 G2
|
|
* Monitor: old Dell TFT display
|
|
* Keyboard: generic USB keyboard
|
|
|
|
No other peripherals or cables were connected. I placed the compute node in my
|
|
cellar (~6ft below ground level) and I remained with the node during the entire
|
|
time it was computing, without using any other devices in the vicinity (no
|
|
mobile phone etc.) The only cables coming out of the box were the two power
|
|
cables, one for the PSU and one for the monitor.
|
|
|
|
Image: <A HREF="https://pbs.twimg.com/media/DOT55KUXUAEV44-.jpg:large">https://pbs.twimg.com/media/DOT55KUXUAEV44-.jpg:large</A>
|
|
|
|
Procedure
|
|
=========
|
|
|
|
I booted the node, with "Try Ubuntu" (Live CD mode). Then, I inserted the
|
|
challenge USB stick and ran `./compute` in the USB media directory, entering
|
|
some additional entropy as requested by typing randomly on the keyboard. The
|
|
box lid was only partially opened to allow use of the keyboard and to view the
|
|
monitor at this point. After 60 minutes had passed, I looked inside the lid
|
|
and saw that the computation had completed, so I wrote down the BLAKE2b hash,
|
|
and unmounted and removed the USB stick, and then powered the node down.
|
|
|
|
Postprocessing
|
|
==============
|
|
|
|
I took the USB stick and transferred the response file to my laptop, and then
|
|
uploaded it using the laptop to S3 via Sean Bowe's transcript site.
|
|
|
|
I did not destroy the compute node but I'm unlikely to use it or plug it in for
|
|
some time.
|
|
--
|
|
Jason Davies, <A HREF="https://www.jasondavies.com">https://www.jasondavies.com</A>
|
|
|
|
Type: application/octet-stream
|
|
Size: 3151 bytes
|
|
Desc: not available
|
|
URL: <<A HREF="/pipermail/zapps-wg/attachments/20171111/1fa7707d/attachment.obj">/pipermail/zapps-wg/attachments/20171111/1fa7707d/attachment.obj</A>>
|
|
><i> On 10 Nov 2017, at 22:11, Sean Bowe via zapps-wg <<A HREF="/mailman/listinfo/zapps-wg">zapps-wg at lists.z.cash.foundation</A>> wrote:
|
|
</I>><i>
|
|
</I>><i> Thanks Andrew! That's a great start.
|
|
</I>><i>
|
|
</I>><i> Now it's Jason Davies' turn.
|
|
</I>><i>
|
|
</I>><i> The entire transcript will appear here throughout the process:
|
|
</I>><i>
|
|
</I>><i> <A HREF="https://powersoftau-transcript.s3-us-west-2.amazonaws.com/index.html">https://powersoftau-transcript.s3-us-west-2.amazonaws.com/index.html</A>
|
|
</I>><i>
|
|
</I>><i> We can make a more formal announcement once we're in the groove and
|
|
</I>><i> everything looks good. We're getting a repo up with attestations soon
|
|
</I>><i> also.
|
|
</I>><i>
|
|
</I>><i> Sean
|
|
</I>><i>
|
|
</I>><i> On Fri, Nov 10, 2017 at 12:53 PM, Andrew Miller <<A HREF="/mailman/listinfo/zapps-wg">soc1024 at illinois.edu</A>> wrote:
|
|
</I>>><i> OK, I'll go first. Below is my report:
|
|
</I>>><i>
|
|
</I>>><i> Powers of Tau Operational writeup
|
|
</I>>><i> =================================
|
|
</I>>><i> Round: 1
|
|
</I>>><i> Date: 2011-11-10
|
|
</I>>><i> Name: Andrew Miller
|
|
</I>>><i> Location: Champaign, Illinois
|
|
</I>>><i>
|
|
</I>>><i> Challenge: (genesis)
|
|
</I>>><i> ce00f2100dd876fdff8dd824f55307bcb72d724f29ff20b9e0760f3a65e5588a65eaed57cbc61697111ae1f4cc7da2e62a85311c2ae683a041fb872b891c68dc
|
|
</I>>><i> Response:
|
|
</I>>><i> 15729e0edc4201dc5ee6241437d926f614cb4214ff1b9c6fbd73daf401639f7a4238cf04bc94edac9f2ad037003daab9a4408ba7c62a4413dc2a0ddd683bd719
|
|
</I>>><i> ./response-2017-11-10-amiller
|
|
</I>>><i>
|
|
</I>>><i> Preparation steps
|
|
</I>>><i> =================
|
|
</I>>><i> I used Sean?s powersoftau rust repo, commit
|
|
</I>>><i> 9e1553c437183540392a7231d0788318a19b18a3
|
|
</I>>><i>
|
|
</I>>><i> I followed instructions online for building portable rust binaries,
|
|
</I>>><i> and so I ran
|
|
</I>>><i> ```
|
|
</I>>><i> cargo build --target=x86_64-unknown-linux-musl --release
|
|
</I>>><i> --features=u128-support --bin=compute
|
|
</I>>><i> ```
|
|
</I>>><i>
|
|
</I>>><i> Compiler: rustc 1.23.0-nightly (02004ef78 2017-11-08)
|
|
</I>>><i>
|
|
</I>>><i> I copied the resulting binary to a freshly formatted USB stick I had.
|
|
</I>>><i>
|
|
</I>>><i> b2sum:
|
|
</I>>><i> 9059a0a64f5021c36df630ca48ac40674862b2fea14f4843ff2150256b95162ac4d6d1621d2dd3f5d0d1c604ad8e581c0ff449d2449140380eab075a9b83c960
|
|
</I>>><i> ./target/x86_64-unknown-linux-musl/release/compute
|
|
</I>>><i>
|
|
</I>>><i> I also rummaged through my shelf of several USB sticks, and found one
|
|
</I>>><i> that happened to be a Linux Mint 18 USB bootable disk, so I used that
|
|
</I>>><i> for my operating system.
|
|
</I>>><i>
|
|
</I>>><i> Sidechannel defenses
|
|
</I>>><i> ====================
|
|
</I>>><i> I used an airgap compute node, a Dell Inspiron that I?ve had for about
|
|
</I>>><i> a year now (Actually this is a computer I bought last year for
|
|
</I>>><i> dress-rehearsals in the Zcash Sprout param generation ceremony).
|
|
</I>>><i>
|
|
</I>>><i> I unplugged all the computer?s hard drives, and detached its
|
|
</I>>><i> wifi/bluetooth radios. I booted the computer from the Linux Mint
|
|
</I>>><i> livecd usb stick, and then also copied the binaries into RAM. The
|
|
</I>>><i> compute node was located in my bedroom, and I attended it for the ~1hr
|
|
</I>>><i> duration of the compute process.
|
|
</I>>><i>
|
|
</I>>><i> Image: <A HREF="https://pbs.twimg.com/media/DOSZz4FXkAEKC7N.jpg:large">https://pbs.twimg.com/media/DOSZz4FXkAEKC7N.jpg:large</A>
|
|
</I>>><i>
|
|
</I>>><i> Postprocessing
|
|
</I>>><i> ==============
|
|
</I>>><i> After compute was finished, I took a cell phone picture of the blake2b
|
|
</I>>><i> hash of the response. I then copied the response file to the USB stick
|
|
</I>>><i> containing the binaries, and then I unplugged the compute node. Using
|
|
</I>>><i> my personal laptop, I posted the blake2b hash to the #mpc chat and
|
|
</I>>><i> uploaded the response file to s3.
|
|
</I>>><i>
|
|
</I>>><i> The repsonse file is hosted here for now, though I expect we'll
|
|
</I>>><i> mirror it elsewhere later:
|
|
</I>>><i> <A HREF="https://s3.amazonaws.com/socrates1024_a/response-2017-11-10-amiller">https://s3.amazonaws.com/socrates1024_a/response-2017-11-10-amiller</A>
|
|
</I>>><i>
|
|
</I>>><i> I did not destroy the compute node and do plan to use it again,
|
|
</I>>><i> although I'm going to leave it unplugged for several days.
|
|
</I>>><i>
|
|
</I>>><i> On Wed, Nov 8, 2017 at 10:19 PM, Sean Bowe <<A HREF="/mailman/listinfo/zapps-wg">sean at z.cash</A>> wrote:
|
|
</I>>>><i> Note that the `response` file contains a hash of the `challenge` file
|
|
</I>>>><i> that was used as input for the compute tool. As a result, only the
|
|
</I>>>><i> hashes of the `response` files need to be published; a hash chain is
|
|
</I>>>><i> formed through all participants. The initial challenge file is
|
|
</I>>>><i> deterministic. (You can use the `new` tool on the repository to
|
|
</I>>>><i> construct it.)
|
|
</I>>>><i>
|
|
</I>>>><i> The initial challenge file has BLAKE2b hash:
|
|
</I>>>><i>
|
|
</I>>>><i> ce00f2100dd876fdff8dd824f55307bcb72d724f29ff20b9e0760f3a65e5588a65eaed57cbc61697111ae1f4cc7da2e62a85311c2ae683a041fb872b891c68dc
|
|
</I>>>><i>
|
|
</I>>>><i> It doesn't hurt to post hashes of everything though. Hash all the things.
|
|
</I>>>><i>
|
|
</I>>>><i> Sean
|
|
</I>>>><i>
|
|
</I>>>><i> On Wed, Nov 8, 2017 at 4:51 PM, Andrew Miller <<A HREF="/mailman/listinfo/zapps-wg">soc1024 at illinois.edu</A>> wrote:
|
|
</I>>>>><i> Thanks Sean!
|
|
</I>>>>><i>
|
|
</I>>>>><i> My idea is to use an ad hoc and publicly visible process. "Get in
|
|
</I>>>>><i> contact with [sean]" could be as simple as posting in public to this
|
|
</I>>>>><i> thread. Unless we're overrun by trolls, a public mailing list can be
|
|
</I>>>>><i> an informal way to agree on who goes next. Whoever posts and says "Me,
|
|
</I>>>>><i> me! I'd like to go next", should, by convention, go next. Any
|
|
</I>>>>><i> aberrations (parties taking too long or dropping out, posting invalid
|
|
</I>>>>><i> data, etc., can be dealt with as needed).
|
|
</I>>>>><i>
|
|
</I>>>>><i> I believe it's also the case that
|
|
</I>>>>><i> a) The "response" file from each person is roughly the same as the
|
|
</I>>>>><i> "challenge" file for the next participant, and
|
|
</I>>>>><i> b) The response/challenge files are safe to be published at any time,
|
|
</I>>>>><i> not private at all.
|
|
</I>>>>><i> So, by convention, we should post the hashes of those files here right
|
|
</I>>>>><i> away, and make a best effort to mirror them publicly (each one is like
|
|
</I>>>>><i> a gigabyte, I think).
|
|
</I>>>>><i>
|
|
</I>>>>><i> What does the initial challenge file consist of? Could you post the
|
|
</I>>>>><i> hash of it here?
|
|
</I>>>>><i>
|
|
</I>>>>><i> Cheers,
|
|
</I>>>>><i>
|
|
</I>>>>><i> On Wed, Nov 8, 2017 at 3:04 PM, Sean Bowe via zapps-wg
|
|
</I>>>>><i> <<A HREF="/mailman/listinfo/zapps-wg">zapps-wg at lists.z.cash.foundation</A>> wrote:
|
|
</I>>>>>><i> Ariel Gabizon, Ian Miers and I have just published a new paper detailing a
|
|
</I>>>>>><i> multi-party computation (MPC) protocol for constructing zk-SNARK public
|
|
</I>>>>>><i> parameters.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> <A HREF="https://eprint.iacr.org/2017/1050">https://eprint.iacr.org/2017/1050</A>
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> The highlights are:
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> * It allows for a single, gigantic ceremony to take place for all possible
|
|
</I>>>>>><i> zk-SNARK circuits within a given size bound. The results of this ceremony
|
|
</I>>>>>><i> are partial zk-SNARK parameters for the entire community. We call this
|
|
</I>>>>>><i> communal ceremony the Powers of Tau.
|
|
</I>>>>>><i> * If you want to use zk-SNARKs in your protocols, you still have to do an
|
|
</I>>>>>><i> MPC for your circuit. But because of the Powers of Tau ceremony, your
|
|
</I>>>>>><i> ceremony is much cheaper to perform and the costs per-participant scale
|
|
</I>>>>>><i> linearly with respect to the circuit complexity.
|
|
</I>>>>>><i> * The best part is that the Powers of Tau and these circuit-specific MPCs
|
|
</I>>>>>><i> can scale to hundreds/thousands of participants. As the number of
|
|
</I>>>>>><i> participants grows, it becomes unrealistic that all of them could be
|
|
</I>>>>>><i> compromised.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> So, let's do the Powers of Tau ceremony! The Zcash Foundation is excited to
|
|
</I>>>>>><i> participate in the process. The Zcash Company is particularly excited in
|
|
</I>>>>>><i> starting soon because we want to leverage it for our next MPC for the
|
|
</I>>>>>><i> Sapling upgrade of Zcash.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> The MPC protocol for this ceremony only requires that one participant
|
|
</I>>>>>><i> successfully destroy the secret randomness they sample during their part. We
|
|
</I>>>>>><i> intend to give participants total flexibility in deciding how to
|
|
</I>>>>>><i> participate; we don't mind what software, hardware or OS you use.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> I have written some Rust software for participants to run:
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> <A HREF="https://github.com/ebfull/powersoftau">https://github.com/ebfull/powersoftau</A>
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> In order to simplify auditing, I won't be making any more changes to the
|
|
</I>>>>>><i> code unless absolutely necessary. You don't have to use this software, but
|
|
</I>>>>>><i> there are no alternative implementations at this time. I think it should be
|
|
</I>>>>>><i> feasible to write a C version of the code using the RELIC toolkit, which has
|
|
</I>>>>>><i> implemented BLS12-381. I am very confident in the Rust code, though, and I
|
|
</I>>>>>><i> believe in its stability/correctness.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> I have some opinions about the ceremony:
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> 1. I disagree with processes that don't improve security of the ceremony.
|
|
</I>>>>>><i> Having a small surface area of code and process increases the chance that
|
|
</I>>>>>><i> bugs will be discovered by auditors because there are fewer things that can
|
|
</I>>>>>><i> go wrong. Remember that there is already quite a bit for the public to
|
|
</I>>>>>><i> check: the transcript correctness, the code correctness, the randomness
|
|
</I>>>>>><i> beacon, the cryptographic proof, code dependencies, etc.
|
|
</I>>>>>><i> 2. It needs to start soon so that it can be useful for the Sapling MPC.
|
|
</I>>>>>><i> 3. It needs to have lots of reputable participants by the time we start the
|
|
</I>>>>>><i> Sapling MPC.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> Given the above, I would like to suggest that we start the ceremony now
|
|
</I>>>>>><i> using my existing code, which supports circuits up to 2^21 gates. This means
|
|
</I>>>>>><i> people would just get in contact with me if they want to participate and
|
|
</I>>>>>><i> I'll schedule them in. I'll try to prioritize reputable people, but I'll
|
|
</I>>>>>><i> allow pretty much anyone I have time to. Everything that I do is publicly
|
|
</I>>>>>><i> verifiable (there is a transcript at the end of the ceremony which people
|
|
</I>>>>>><i> can check).
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> Andrew Miller has a few interesting ideas for a more distributed process for
|
|
</I>>>>>><i> scheduling "who goes next" but there are some disadvantages and risks
|
|
</I>>>>>><i> involved IMO. In any case, the process can be changed later without
|
|
</I>>>>>><i> affecting anything, so I don't see a purpose in delaying the start of the
|
|
</I>>>>>><i> ceremony on such things.
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> I'd like to hear from others about this plan so we can begin soon!
|
|
</I>>>>>><i>
|
|
</I>>>>>><i> Sean Bowe
|
|
</I>>>>>><i> Zcash Company
|
|
</I>>>>><i>
|
|
</I>>>>><i>
|
|
</I>>>>><i>
|
|
</I>>>>><i> --
|
|
</I>>>>><i> Andrew Miller
|
|
</I>>>>><i> University of Illinois at Urbana-Champaign
|
|
</I>>><i>
|
|
</I>>><i>
|
|
</I>>><i>
|
|
</I>>><i> --
|
|
</I>>><i> Andrew Miller
|
|
</I>>><i> University of Illinois at Urbana-Champaign
|
|
</I>
|
|
</PRE>
|
|
|
|
<!--endarticle-->
|
|
<HR>
|
|
<P><UL>
|
|
<!--threads-->
|
|
<LI>Previous message (by thread): <A HREF="000008.html">[zapps-wg] Powers of Tau Ceremony Proposal
|
|
</A></li>
|
|
<LI>Next message (by thread): <A HREF="000010.html">[zapps-wg] Powers of Tau Ceremony Proposal
|
|
</A></li>
|
|
<LI> <B>Messages sorted by:</B>
|
|
<a href="date.html#9">[ date ]</a>
|
|
<a href="thread.html#9">[ thread ]</a>
|
|
<a href="subject.html#9">[ subject ]</a>
|
|
<a href="author.html#9">[ author ]</a>
|
|
</LI>
|
|
</UL>
|
|
|
|
<hr>
|
|
<a href="/mailman/listinfo/zapps-wg">More information about the zapps-wg
|
|
mailing list</a><br>
|
|
</body></html>
|