From: Rusty Russell <rusty@rustcorp.com.au>
To: Gregory Maxwell <greg@xiph.org>,
	Bitcoin Protocol Discussion <bitcoin-dev@lists.linuxfoundation.org>
In-Reply-To: <CAAS2fgRePSQ=-3MTR3p3U1zbd1ucfNg0_ocAegJCi4qR=XpypA@mail.gmail.com>
References: <5727D102.1020807@mattcorallo.com> <5730C37E.2000004@gmail.com>
	<87twi6zdl8.fsf@rustcorp.com.au>
	<CAAS2fgRePSQ=-3MTR3p3U1zbd1ucfNg0_ocAegJCi4qR=XpypA@mail.gmail.com>
User-Agent: Notmuch/0.21 (http://notmuchmail.org) Emacs/24.5.1
	(x86_64-pc-linux-gnu)
Date: Wed, 11 May 2016 06:53:55 +0930
Message-ID: <87k2j1zjx0.fsf@rustcorp.com.au>
MIME-Version: 1.0
Content-Type: text/plain
Subject: Re: [bitcoin-dev] Compact Block Relay BIP
Precedence: list

Gregory Maxwell <greg@xiph.org> writes:
> On Tue, May 10, 2016 at 5:28 AM, Rusty Russell via bitcoin-dev
> <bitcoin-dev@lists.linuxfoundation.org> wrote:
>> I used variable-length bit encodings, and used the shortest encoding
>> which is unique to you (including mempool).  It's a little more work,
>> but for an average node transmitting a block with 1300 txs and another
>> ~3000 in the mempool, you expect about 12 bits per transaction.  IOW,
>> about 1/5 of your current size.  Critically, we might be able to fit in
>> two or three TCP packets.
>
> Hm. 12 bits sounds very small even giving those figures. Why failure
> rate were you targeting?

That's a good question; I was assuming a best-case in which we have
mempool set reconciliation (handwave) thus know they are close.  But
there's also an alterior motive: any later more sophisticated approach
will want variable-length IDs, and I'd like Matt to do the work :)

In particular, you can significantly narrow the possibilities for a
block by sending the min-fee-per-kb and a list of "txs in my mempool
which didn't get in" and "txs which did despite not making the
fee-per-kb".  Those turn out to be tiny, and often make set
reconciliation trivial.  That's best done with variable-length IDs.

> (*Not interesting because it mostly reduces exposure to loss and the
> gods of TCP, but since those are the long poles in the latency tent,
> it's best to escape them entirely, see Matt's udp_wip branch.)

I'm not convinced on UDP; it always looks impressive, but then ends up
reimplementing TCP in practice.  We should be well within a TCP window
for these, so it's hard to see where we'd win.

>> I would also avoid the nonce to save recalculating for each node, and
>> instead define an id as:
>
> Doing this would greatly increase the cost of a collision though, as
> it would happen in many places in the network at once over the on the
> network at once, rather than just happening on a single link, thus
> hardly impacting overall propagation.

"Greatly increase"?  I don't see that.

Let's assume an attacker grinds out 10,000 txs with 128 bits of the same
TXID, and gets them all in a block.  They then win the lottery and get a
collision.  Now we have to transmit ~48 bytes more than expected.

> Using the same nonce means you also would not get a recovery gain from
> jointly decoding using compact blocks sent from multiple peers (which
> you'll have anyways in high bandwidth mode).

Not quite true, since if their mempools differ they'll use different
encoding lengths, but yes, you'll get less of this.

> With a nonce a sender does have the option of reusing what they got--
> but the actual encoding cost is negligible, for a 2500 transaction
> block its 27 microseconds (once per block, shared across all peers)
> using Pieter's suggestion of siphash 1-3 instead of the cheaper
> construct in the current draft.
>
> Of course, if you're going to check your whole mempool to reroll the
> nonce, thats another matter-- but that seems wasteful compared to just
> using a table driven size with a known negligible failure rate.

I'm not worried about the sender: The recipient needs to encode all the
mempool.

>> As Peter R points out, we could later enhance receiver to brute force
>> collisions (you could speed that by sending a XOR of all the txids, but
>> really if there are more than a few collisions, give up).
>
> The band between "no collisions" and "infeasible many" is fairly
> narrow.  You can add a small amount more space to the ids and
> immediately be in the no collision zone.

Indeed, I would be adding extra bits in the sender and not implementing
brute force in the receiver.  But I welcome someone else to do so.

Cheers,
Rusty.