bitpro가 원문에서 글자만 모은 내용이에요. 표·그림·수식은 빠지거나 깨질 수 있어요. 전체 7쪽 중 2쪽
o the Yellow Paper Council [24]. As
the ultimate authority, the underlying stakable token hold-
ers would have “referendum” control. To reflect the users’
need for development but the developers’ need for legiti-
macy, we expect a reasonable direction would be to form
the two chambers from a “user” committee (made up of
bonded validators) and a “technical” committee made up
of major client developers and ecosystem players. The
body of token holders would maintain the ultimate legit-
imacy and form a supermajority to augment, reparam-
eterise, replace or dissolve this structure, something we
don’t doubt the eventual need for: in the words of Twain
“Governments and diapers must be changed often, and for
the same reason ”.
Whereas reparameterisation is typically trivial to ar-
range within a larger consensus mechanism, more qualita-
tive changes such as replacement and augmentation would
likely need to be either non-automated “soft-decrees” (e.g.
through the canonicalisation of a block number and the
hash of a document formally specifying the new protocol)
or necessitate the core consensus mechanism to contain a
sufficiently rich language to describe any aspect of itself
which may need to change. The latter is an eventual aim,
however, the former more likely to be chosen in order to
facilitate a reasonable development timeline.
Polkadot’s primary tenets and the rules within which
we evaluate all design decisions are:
Minimal: Polkadot should have as little functional-
ity as possible.
Simple: no additional complexity should be present
in the base protocol than can reasonably be
offloaded into middleware, placed through a
parachain or introduced in a later optimisation.
General: no unnecessary requirement, constraint
or limitation should be placed on parachains;
Polkadot should be a test bed for consensus sys-
tem development which can be optimised through
making the model into which extensions fit as ab-
stract as possible.
Robust: Polkadot should provide a fundamentally
stable base-layer. In addition to economic sound-
ness, this also means decentralising to minimise
the vectors for high-reward attacks.
4. Participation in Polkadot
There are four basic roles in the upkeep of an Polkadot
network: collator, fisherman, nominator and validator. In
one possible implementation of Polkadot, the latter role
may actually be broken down into two roles: basic valida-
tor and availability guarantor; this is discussed in section
6.5.3.
Collator Fisherman
Validators
(this group)
Validators
(other groups)
approves
becomes
monitors
reports
bad
behaviour to
provides block
candidates
for
Nominator
Figure 1. The interaction between the
four roles of Polkadot.
4.1. Validators. A validator is the highest charge and
helps seal new blocks on the Polkadot network. The val-
idator’s role is contingent upon a sufficiently high bond
being deposited, though we allow other bonded parties to
nominate one or more validators to act for them and as
such some portion of the validator’s bond may not neces-
sarily be owned by the validator itself but rather by these
nominators.
A validator must run a relay-chain client implementa-
tion with high availability and bandwidth. At each block
the node must be ready to accept the role of ratifying
a new block on a nominated parachain. This process
involves receiving, validating and republishing candidate
blocks. The nomination is deterministic but virtually un-
predictable much in advance. Since the validator cannot
reasonably be expected to maintain a fully-synchronised
database of all parachains, it is expected that the valida-
tor will nominate the task of devising a suggested new
parachain block to a third-party, known as a collator.
Once all new parachain blocks have been properly rat-
ified by their appointed validator subgroups, validators
must then ratify the relay-chain block itself. This involves
updating the state of the transaction queues (essentially
moving data from a parachain’s output queue to another
parachain’s input queue), processing the transactions of
the ratified relay-chain transaction set and ratifying the
final block, including the final parachain changes.
POLKADOT: VISION FOR A HETEROGENEOUS MUL TI-CHAIN FRAMEWORK DRAFT 1 5
A validator not fulfilling their duty to find consensus
under the rules of our chosen consensus algorithm is pun-
ished. For initial, unintentional failures, this is through
withholding the validator’s reward. Repeated failures re-
sult in the reduction of their security bond (through burn-
ing). Provably malicious actions such as double-signing or
conspiring to provide an invalid block result in the loss of
the entire bond (which is partially burnt but mostly given
to the informant and the honest actors).
In some sense, validators are similar to the mining pools
of current PoW blockchains.
4.2. Nominators. A nominator is a stake-holding party
who contributes to the security bond of a validator. They
have no additional role except to place risk capital and as
such to signal that they trust a particular validator (or
set thereof) to act responsibly in their maintenance of the
network. They receive a pro-rata increase or reduction
in their deposit according to the bond’s growth to which
they contribute.
Together with collators, next, nominators are in some
sense similar to the miners of the present-day PoW net-
works.
4.3. Collators. Transaction collators (collators for short)
are parties who assist validators in producing valid
parachain blocks. They maintain a “full-node” for a par-
ticular parachain; meaning that they retain all necessary
information to be able to author new blocks and execute
transactions in much the same way as miners do on cur-
rent PoW blockchains. Under normal circumstances, they
will collate and execute transactions to create an unsealed
block, and provide it, together with a zero-knowledge
proof, to one or more validators presently responsible for
proposing a parachain block.
The precise nature of the relationship between colla-
tors, nominators and validators will likely change over
time. Initially, we expect collators to work very closely
with validators, since there will be only a few (perhaps
only one) parachain(s) with little transaction volume. The
initial client implementation will include RPCs to allow a
parachain collator node to unconditionally supply a (relay-
chain) validator node with a provably valid parachain
block. As the cost of maintaining a synced version of
all such parachains increases, we expect to see additional
infrastructure in place which will help separate out the
duties to independent, economically-motivated, parties.
Eventually, we expect to see collator pools who vie to
collect the most transaction fees. Such collators may be-
come contracted to serve particular validators over a pe-
riod of time for an on-going share in the reward proceeds.
Alternatively, “freelance” collators may simply create a
market offering valid parachain blocks in return for a com-
petitive share of the reward payable immediately. Simi-
larly, decentralised nominator pools would allow multiple
bonded participants to coordinate and share the duty of a
validator. This ability to pool ensures open participation
leading to a more decentralised system.
4.4. Fishermen. Unlike the other two active parties,
fishermen are not directly related to the block-authoring
process. Rather they are independent “bounty hunters”
motivated by a large one-off reward. Precisely due to
the existence of fishermen, we expect events of misbe-
haviour to happen seldom, and when they do only due to
the bonded party being careless with secret key security,
rather than through malicious intent. The name comes
from the expected frequency of reward, the minimal re-
quirements to take part and the eventual reward size.
Fishermen get their reward through a timely proof that
at least one bonded party acted illegally. Illegal actions
include signing two blocks each with the same ratified par-
ent or, in the case of parachains, helping ratify an invalid
block. To prevent over-rewarding or the compromise and
illicit use of a session’s secret key, the base reward for
providing a single validator’s illegally signed message is
minimal. This reward increases asymptotically as more
corroborating illegal signatures from other validators are
provided implying a genuine attack. The asymptote is set
at 66% following our base security assertion that at least
two-thirds of the validators act benevolently.
Fishermen are somewhat similar to “full nodes” in
present-day blockchain systems that the resources needed
are relatively small and the commitment of stable uptime
and bandwidth is not necessary. Fishermen differ in so
much as they must post a small bond. This bond prevents
sybil attacks from wasting validators’ time and compute
resources. It is immediately withdrawable, probably no
more than the equivalent of a few dollars and may lead
to reaping a hefty reward from spotting a misbehaving
validator.
5. Design Overview
This section is intended to give a brief overview of the
system as a whole. A more thorough exploration of the
system is given in the section following it.
5.1. Consensus. On the relay-chain, Polkadot achieves
low-level consensus over a set of mutually agreed valid
blocks through a modern asynchronous Byzantine fault-
tolerant (BFT) algorithm. The algorithm will be inspired
by the simple Tendermint [11] and the substantially more
involved HoneyBadgerBFT [14]. The latter provides an
efficient and fault-tolerant consensus over an arbitrarily
defective network infrastructure, given a set of mostly be-
nign authorities or validators.
For a proof-of-authority (PoA) style network, this alone
would be sufficient, however Polkadot is imagined to be
also deployable as a network in a fully open and public
situation without any particular organisation or trusted
authority required to maintain it. As such we need a
means of determining a set of validators and incentivising
them to be honest. For this we utilise PoS based selection
criteria.
5.2. Proving the Stake. We assume that the network
will have some means of measuring how much “stake”
any particular account has. For ease of comparison to
pre-existing systems, we will call the unit of measurement
“tokens”. Unfortunately the term is less than ideal for a
number of reasons, not least that being simply a scalar
value associated with an account, there is no notion of
individuality.
We imagine validators be elected, infrequently (at most
once per day but perhaps as seldom as once per quarter),
through a Nominated Proof-of-Stake (NPoS) scheme. In-
centivisation can happen through a pro-rata allocation of
POLKADOT: VISION FOR A HETEROGENEOUS MUL TI-CHAIN FRAMEWORK DRAFT 1 6
Relay
chain
Validator swarm
(each coloured by its
designated parachain)
Transaction
(submitted by
external actor)
Parachain
bridge
Virtual parachain
(e.g. Ethereum)
Parachain Parachain
queues and I/O
Propagated transactions
Block candidate submission
2nd order
Relay-chain
Parachain community
Account
Inbound transaction
Outbound transaction
Interchain transactions
(managed by validators)
Collator
Propagated block
Fisherman
Figure 2. A summary schematic of the Polkadot system. This shows collators collecting and propa-
gating user-transactions, as well as propagating block candidates to fishermen and validators. It also
shows how an account can post a transaction which is carried out of its parachain, via the relay-chain
and on into another parachain where it can be interpreted as a transaction to an account there.
funds coming from a token base expansion (up to 100%
per year, though more likely around 10%) together with
any transaction fees collected. While monetary base ex-
pansion typically leads to inflation, since all token owners
would have a fair opportunity at participation, no token-
holder would need to suffer a reduction in value of their
holdings over time provided they were happy to take a
role in the consensus mechanism. A particular proportion
of tokens would be targeted for the staking process; the
effective token base expansion would be adjusted through
a market-based mechanism to reach this target.
Validators are bonded heavily by their stakes; exiting
validators’ bonds remain in place long after the valida-
tors’ duties cease (perhaps around 3 months). This long
bond-liquidation period allows future misbehaviour to be
punished up until the periodic checkpointing of the chain.
Misbehaviour results in punishment, such as reduction of
reward or, in cases which intentionally compromise the
network’s integrity, the validator losing some or all of its
stake to other validators, informants or the stakeholders
as a whole (through burning). For example, a validator
who attempts to ratify both branches of a fork (sometimes
known as a “short-range” attack) may be identified and
punished in the latter way.
Long-range “nothing-at-stake” attacks 4 are circum-
vented through a simple “checkpoint” latch which pre-
vents a dangerous chain-reorganisation of more than a
particular chain-depth. To ensure newly-syncing clients
are not able to be fooled onto the wrong chain, regular
“hard forks” will occur (of at most the same period of the
validators’ bond liquidation) that hard-code recent check-
point block hashes into clients. This plays well with a fur-
ther footprint-reducing measure of “finite chain length” or
periodic reseting of the genesis-block.
5.3. Parachains and Collators. Each parachain gets
similar security affordances to the relay-chain: the
parachains’ headers are sealed within the relay-chain block
ensuring no reorganisation, or “double-spending”, is possi-
ble following confirmation. This is a similar security guar-
antee to that offered by Bitcoin’s side-chains and merge-
mining. Polkadot, however, also provides strong guaran-
tees that the parachains’ state transitions are valid. This
happens through the set of validators being cryptograph-
ically randomly segmented into subsets; one subset per
parachain, the subsets potentially differing per block. This
setup generally implies that parachains’ block times will
be at least as long as that of the relay-chain. The specific
means of determining the partitioning is outside the scope
4Such an attack is where the adversary forges an entirely new chain of history from the genesis block onwards. Through controlling a
relatively insignificant portion of stake at the offset, they are able to incrementally increase their portion of the stake relative to all other
stakeholders as they are the only active participants in their alternative history. Since no intrinsic physical limitation exists on the creation
of blocks (unlike PoW where quite real computational energy must be spent), they are able to craft a chain longer than the real chain in a
relatively short timespan and potentially make it the longest and best, taking over the canonical state of the network.
POLKADOT: VISION FOR A HETEROGENEOUS MUL TI-CHAIN FRAMEWORK DRAFT 1 7
of this document but would likely be based either around
a commit-reveal framework similar to the RanDAO [19] or
use data combined from previous blocks of each parachain
under a cryptographically secure hash.
Such subsets of validators are required to provide a
parachain block candidate which is guaranteed valid (on
pain of bond confiscation). Validity revolves around two
important points; firstly that it is intrinsically valid—that
all state transitions were executed faithfully and that all
external data referenced (i.e. transactions) is valid for in-
clusion. Secondly, that any data which is extrinsic to its
candidate, such as those external transactions, has suffi-
ciently high availability so that participants are able to
download it and execute the block manually. 5 Valida-
tors may provide only a “null” block containing no ex-
ternal “transactions” data, but may run the risk of get-
ting a reduced reward if they do. They work alongside
a parachain gossip protocol with collators—individuals
who collate transactions into blocks and provide a non-
interactive, zero-knowledge proof that the block consti-
tutes a valid child of its parent (and taking any transaction
fees for their trouble).
It is left to parachain protocols to specify their own
means of spam-prevention: there is no fundamental no-
tion of “compute-resource metering” or “transaction fee”
imposed by the relay-chain. There is also no direct en-
forcement on this by the relay-chain protocol (though it
is unlikely that the stakeholders would choose to adopt
a parachain which didn’t provide a decent mechanism).
This is an explicit nod to the possibility of chains unlike
Ethereum, e.g. a Bitcoin-like chain which has a much sim-
pler fee model or some other, yet-to-be-proposed spam-
prevention model.
Polkadot’s relay-chain itself will probably exist as an
Ethereum-like accounts and state chain, possibly an EVM-
derivative. Since the relay-chain nodes will be required to
do substantial other processing, transaction throughput
will be minimised partly through large transaction fees
and, should our research models require, a block size limit.
5.4. Interchain Communication. The critical final in-
gredient of Polkadot is interchain communication. Since
parachains can have some sort of information channel be-
tween them, we allow ourselves to consider Polkadot a
scalable multi-chain. In the case of Polkadot, the commu-
nication is as simple as can be: transactions executing in a
parachain are (according to the logic of that chain) able to
effect the dispatch of a transaction into a second parachain
or, potentially, the relay-chain. Like external transactions
on production blockchains, they are fully asynchronous
and there is no intrinsic ability for them to return any
kind of information back to its origin.
Destination: gets
data from prior
block’s validators.
Account receives post:
entry removed from
ingress Merkle tree
Account sends post:
entry placed in
egress Merkle tree
for destination
parachain
egressSource: sharesdata with next block’s
validators
proof-of-post stored in
parachain egress Merkle
tree
routed reference placed
in destination parachain’s
ingress Merkle tree
ingress
Figure 3. A basic schematic showing
the main parts of routing for posted
transactions (”posts”).
To ensure minimal implementation complexity, min-
imal risk and minimal straight-jacketing of future
parachain architectures, these interchain transactions are
effectively indistinguishable from standard externally-
signed transactions. The transaction has an origin seg-
ment, providing the ability to identify a parachain, and
an address which may be of arbitrary size. Unlike com-
mon current systems such as Bitcoin and Ethereum, in-
terchain transactions do not come with any kind of “pay-
ment” of fee associated; any such payment must be man-
aged through negotiation logic on the source and desti-
nation parachains. A system such as that proposed for
Ethereum’s Serenity release [7] would be a simple means
of managing such a cross-chain resource payment, though
we assume others may come to the fore in due course.
Interchain transactions are resolved using a simple
queuing mechanism based around a Merkle tree to ensure
fidelity. It is the task of the relay-chain maintainers to
move transactions on the output queue of one parachain
into the input queue of the destination parachain. The
passed transactions get referenced on the relay-chain, how-
ever are not relay-chain transactions themselves. To pre-
vent a parachain from spamming another parachain with
transactions, for a transaction to be sent, it is required
that the destination’s input queue be not too large at
the time of the end of the previous block. If the input
queue is too large after block processing, then it is con-
sidered “saturated” and no transactions may be routed to
it within subsequent blocks until reduced back below the
limit. These queues are administered on the relay-chain
allowing parachains to determine each other’s sat프로젝트로 돌아가기