bitpro가 원문에서 글자만 모은 내용이에요. 표·그림·수식은 빠지거나 깨질 수 있어요. 전체 3쪽 중 1쪽
Version 1.0
Disclaimer
The website and white paper prepared by Bifrost is for elaboration, description and for
informational purposes only, and is not an offer or a solicitation to buy or sell any securities
or to invest in any financial instruments. The registration on theBifrost platform permits
access to the services provided by the platform as detailed herein. Registration is not intended
to afford the holder any rights in, or claims to, any of the assets ofBifrost or to in any way
share in any profits that the platform may achieve. Interested parties acknowledge agreeing to
the Consent to Use Electronic Records, Privacy Policy and Terms and Conditions. The white
paper or content on this webiste is subject to change at any time without notification. The white
paper and website describes the current plan and vision for theBifrost platform. While we
intend to attempt to realize this vision, please recognize that it is dependent on quite a number
of factors and subject to quite a number of risks. We do not guarantee, represent or warrant any
of the statements in the white paper or website, because they are based on our current beliefs,
expectations and assumptions, about which there can be no assurance due to various anticipated
and unanticipated events that may occur. Please know that we plan to work hard in seeking
to achieve the vision laid out in the white paper and website, but that you cannot rely on any
of it coming true. Blockchain, cryptocurrencies and other aspects of our technology and these
markets are in their infancy and will be subject to many challenges, competition and a changing
environment. We will try to update our community as things grow and change, butBifrostv0.9
undertake no obligation to do so. Interested parties acknowledge that theBifrost platform,
as described herein, may never in fact operate as intended. Parties also acknowledge that all
services and scope of work proposed in this white paper is subject to any licensing required.
The information and graphical content contained in the white paper and website should not
be construed as a guarantee and is subject to change at any time without prior notification.
The information contained herein is intended for familiarization, and should not be utilized or
reproduced in any form in full or part. The white paper has been prepared to the best of our
knowledge and research, however it should not be relied upon for any future actions including
but not limited to financial or investment related decisions. The company, founders, advisors or
affiliates shall not be liable for any losses that arise in any way due to the use of this document
or the contents contained herein. The content, both written and graphic may be historical or
forward looking and therefore should not be relied upon. The content is based on assumptions
and therefore uses words such as ‘expects’, ‘intends’, ‘will’, ‘can’, ‘should’ or similar expressions.
The assumptions drawn in this document are based on past trends and data from third parties
and other sources, which were believed to be reasonable at the time they were made. However,
they still involve unknown risks and uncertainties, as it is impossible to predict anything outside
of our immediate control including economic factors. Individuals and investors are requested to
carefully consider the risks, costs and benefits of acquiring the token through this crowd sale as
opposed to through a third-party exchange, once operational.
1
Contents
1 Introduction 3
1.1 Motivations and problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Approaches for efficient solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.3 Bifrost method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Bifrost System 5
2.1 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Node architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.4 Execution stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.5 Operation stack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3 Recipe for Multi-component DApps 11
3.1 Motivations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 Recipe language system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 Splitting and transpilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4 Linker & BFC System 15
4.1 Interworking via Linkers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.1.1 BFC operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.1.2 States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.1.3 Actions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.2 DApp-wise and Container-wise Linkers . . . . . . . . . . . . . . . . . . . . . . . . 17
4.3 Relaying state of cross variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.4 Linker availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.5 Relaying cross-function calls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.6 Relaying coins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.7 Linker description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5 Container 20
5.1 Multiple Containers for service availability . . . . . . . . . . . . . . . . . . . . . . 20
5.2 Container manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
5.3 Container/Node operator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
6 Bifrost Ecosystem 21
6.1 Bifrost ecosystem overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.2 Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
6.3 Bootstrapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
7 Conclusion 23
2
1 Introduction
1.1 Motivations and problems
We have high expectations with positive evaluations about future innovations that blockchain
will bring. Bitcoin demonstrated the possibility of the functions of money through decentral-
ization, and Ethereum used smart contract with technological advances to create a foundation
on which many applications could actually operate. Smart contracts and decentralized applica-
tions (DApps) enable various services on the blockchain. As DApp services have increased and
diversified, the blockchain ecosystem has matured. Given the superior incentivization structure,
stability, and transparency under the decentralized network, DApps can be a game changer for
the entire software business in the near future.
However, the speed of DApp’s development is significantly below the public’s expectations,
and there are even doubts about the effectiveness of the blockchain technology itself nowadays.
Particularly, limitations are found, for example, Ethereum network with more than 1,000 DApps
is unable to provide effective technical support due to scaling issues. In addition, efforts to solve
all the problems about decentralization, safety, and scalability (called Trilemma) at once in the
protocol level, have led to the phenomenon of “Protocol Fever,” in which the number of protocols
is larger than that of DApps. Given the fact that DApp developer’s first decision is to select
a blockchain network protocol to provide DApp services, the current phenomenon of Protocol
Fever is a major impediment to stable DApp development environment. The presence of various
protocols and diverse DApps deepens mismatches with each other’s needs.
Despite the structural problems of this blockchain technology, it has revolutionized beyond
the national and industrial sectors, and most of them have been commercialized using their own
managed blockchains (or private blockchains). In the field of copyright and logistics in particular,
the private blockchain is pioneering new areas with unique data reliability and performance
improvements. For instance, “IBM Food Trust,” a blockchain-based food tracking network that
was commercialized in October 2018, is a system that utilizes its own managed blockchain by
sharing all details such as food origin and transportation records.
The ecosystem of dynamic DApp environment is essential for the development of blockchain
as an innovative service. As with many other high-tech fields, the DApp field has difficulties that
must be surmounted to construct its ecosystem and establish stable business service platforms.
In this chapter, based on the technical environment of the blockchain described above, we define
the problems that may arise in the DApp development environments, present the prerequisites
for its solutions, and ultimately show Bifrost solutions for the DApps to work.
The following three problems are found in the DApp development environments:
The first problem is “the platform risk.” DApp developers have to choose a blockchain mainnet
protocol (e.g., Ethereum, EOS) to provide their services. Numerous mainnet protocols have been
developed, and many others with different emphasized functions are currently ready to go public.
All the mainnet protocols are unique and have different advantages and disadvantages; however,
3
none provides a perfect development environment for DApps. Nonetheless, developers have to
choose one and accept the uncontrollable risk that comes from the forthcoming projects of the
chosen protocols.
Second, DApps have too many different requirements for blockchain mainnets. Since DApps
provide different services, the system requirements can be varied. However, no single mainnet
protocol is able to accept all demands. For example, crypto games must handle a number of
transactions in their domains and transfer token rewards safely. That is, crypto games need
scalability, stability, and decentralization at the same time, which is almost impossible in a single
blockchain network.
Third, developers can choose their own managed blockchain (e.g., private blockchains), but
the networks have limitations. Many research studies are being carried out to obtain more flexi-
bility and scalability using private blockchain systems as a part of mainnet blockchain networks.
However, private blockchain systems have more difficulties in verifying safety and reliability to
construct DApp ecosystems than externally operating blockchain networks.
1.2 Approaches for efficient solutions
We consider the following two approaches to get efficient and practical solutions to the three
problems outlined above in the DApp development environment.
First, no single blockchain mainnet protocol can be the solution for a DApp platform. Since
we do not have “the perfect blockchain,” it is inefficient and risky to depend on one blockchain
mainnet as a platform.
Second, adding OS or abstraction layers increases overheads only. Implementing improve-
ments via mainnet blockchain extension may overload the entire system and limit scalability.
1.3 Bifrost method
The premises above suggest that we need a brand new platform, where we can use existing
blockchain systems as building blocks to achieve the strength of each system and organize en-
vironments for independent DApp services at the same time. That is, if we can make a Dapp
platform that combines multiple blockchains that provides the reliability and safety as well as
some customizability with using internally operating blockchain network, the problems of the
DApp development environment can be solved.
We introduce “Bifrost,” the new standard DApp platform. Bifrost focuses on DApps’
practical services on blockchains.Bifrostprovides a multiple blockchain platform where we can
compose existing qualified blockchain systems as components and restructure DApps so that they
can run on the combined blockchain systems. Inside of the platform,Bifrostsupports the entire
process, from DApp deployment to DApp operation. In the deployment phase,Bifrost splits
DApps’ codes into pre-designed blockchain parts and restructures them automatically. After the
components are deployed in each position,Bifrost serves as a “Linker,” which communicates
4
with deployed components and supports safe execution in the operation stage.Bifrostoffers an
improved DApp platform, which has the safety and the high performance of multiple blockchain
systems.
Bifrost has flexibility to prepare for the uncertain future of blockchain mainnet protocols.
Bifrost has the following characteristics:
1. Blockchains as building blocks. Bifrost constructs a parallel environment by com-
bining multiple blockchains, which can come together and form a federation.
2. Container applications. One blockchain can form an independent Container and each
Containercanhaveadifferentsetup. Thatis, thecreationanddestructionoftheContainers
are dynamic, and performance tuning, such as increasing the speed of transactions or
enhancing anonymity, can be applied to the Containers. In addition, a new approach
to managing the Containers is possible. For instance, an independent Container can be
operated and destroyed after execution if the service must not be recorded.
3. Multi-component DApp reconstruction. DApps, including smart contracts on the
Bifrost platform are reconstructed to have appropriate components for each blockchain
and run simultaneously. Therefore, DApp can implement the high performance as well
as the safety verification of multiple blockchains. The reconstruction technology of multi-
component DApps is working for the existing DApps, and it will be upgraded to support
the full flexibility ofBifrost by introducing a new high-level programming language.
Bifrost has the real potential to expand the horizons of DApps. DApps can ultimately
overcome platform risk from the blockchain mainnet protocol and utilize the strength of more
than one blockchain by using combinations.
2 Bifrost System
2.1 Design
DApps’ dilemma or trilemma. A number of blockchain systems are being introduced, each
with their own unique features. DApp developers are trying to find a blockchain system that
matches their own requirements perfectly. However, it is not an easy task. Once a DApp is
built on the a certain blockchain system, it can reap the benefits of the blockchain’s advantages
(e.g., high scalability), but it must also tolerate its disadvantages (e.g., limited safety). Consider a
DApp used to sell digital contents via a blockchain system. Some blockchain systems are suitable
for operations for asset transfers or ownership publications. On the contrary, some blockchain
systems are preferred for inside auctions or negotiations. Under this dilemma (or trilemma), a
DApp service provider is faced with the challenge of picking a blockchain system that optimizes
the performance and security of its service.
5
Don’t reinvent the wheel. Most of the newly proposed blockchain systems have a common,
ambitious approach: building a completely new blockchain system that solves all the problems at
once, particularly those related to scalability. However, every DApp has its own goals and needs.
A single blockchain system cannot satisfy all the requirements. Furthermore, no blockchain
system is free from intrinsic and technical tradeoffs. Thus, we are still in a vicious circle, where
a new solution causes another problem.
Our approach: cherry picking in a good way . We present a new DApp platform,Bifrost,
as a practical solution to the problems of DApps. DApp service providers and developers are con-
fronted with two challenges: (1) selecting a mainnet blockchain for their service and (2) meeting
the DApp’s (paradoxical) scalability and security requirements. We tackle the fundamental prob-
lem itself: DApps just need a platform that satisfies all their different requirements. Therefore,
we focus on how DApps can use desirable features from the existing blockchain systems rather
than on making another mainnet blockchain. Bifrost combines multiple blockchain systems
into a unified smart contract environment. It also automatically restructures the DApp itself to
run directly on top of the multiple blockchain system.
Bifrost is a DApp platform that supports a DApp from development to operation. A DApp
service consists of a core smart contract code and interfaces to web, external storage, and users.
A smart contract is represented as a set of variables (state) and functions. Thus, the features of
a DApp can be divided by functions. InBifrost, a DApp developer creates the DApp service
in a conventional way but s/he may also additionally indicate the target blockchain system for
each function in the code. In the deployment phase of the DApp,Bifrost maps the functions
to their targets. It then splits the smart contract code and adjusts (and transpiles) each split
component to its target blockchain system. However, to enable the seamless interoperation of
the split DApp, a platform also needs to support the execution of the DApp as well as the
sophisticated conversion of the smart contract. To do so,Bifrost automatically generates a
description for interfacing the components. In the operation phase, it then bridges them to run
as a single DApp.
Design goals. The design goals ofBifrost are as follows:
• Do not let a DApp be bound to a single blockchain system.
• Provide a unified smart contract environment on top of multiple blockchain systems.
• Let a DApp run directly on the combined environment and reap the benefits from the each
blockchain systems.
• Minimize the overhead of the multiple blockchain DApp. (Provide the performance of
native smart contracts for each blockchain system.)
• Present a new program language to extend the smart contract program model for the
multiple blockchain environments.
6
%FWFMPQNFOU%FQMPZNFOU1IBTFT 0QFSBUJPO1IBTFT
WebOracles
I/O interface
Linker Linker
Ext.
Storage
Secured
(TPM)
3FBMN
#MPDLDIBJO4ZTUFN#
FH
&UIFSFVN
&GGJDJFOU
PSDPOGJEFOUJBM
5Y
1VCMJDMZ
MPHHFE
5Y
%FTJHOB
%"QQ
4FMFDUB#JGSPTU
$POUBJOFSUP
EFQMPZ
%"QQ
JO3FDJQF
%"QQ
DPNQPOFOU
%"QQ
DPNQPOFOU
-JOLFSEFTDSJQUJPO
#MPDLDIBJO4ZTUFN"
#MPDLDIBJOTZTUFNTQFDJGJD
TFUUJOH
FH
QFSGPSNBODFUVOJOH
BFC
Token
System
4QMJU
5SBOTQJMF
3FBMN
#JGSPTU
%"QQ
#JGSPTU
%"QQ
Figure 1: Bifrost process
2.2 Overview
Bifrost provides a DApp platform on the top of the multiple blockchains. In particular, to
achieve the goals in §2.1,Bifrostworks on the lifecycle of the DApp. It supports the DApp from
the development phase to the operation phase. In the conventional integration of heterogeneous
systems, one usually adds an extra abstraction layer, such as a virtual machine or chain, to
establish a unified environment. However, the layered approach causes computation overhead
and restricts the use of subsystem-specific features. Bifrost provides a more fundamental
solution with advanced programming language techniques. A DApp inBifrost is developed for
the multiple blockchain system with little effort, and it is restructured to the native form for
each blockchain system in the deployment phase. For this purpose,Bifrost splits a DApp into
different smart contracts for each system in multiple blockchains, and then transpiles (translates
to a different language) each one to the native smart contract form. Thus, the multiple blockchain
system executes each DApp component directly without overhead. Since Bifrost combines
multiple blockchain systems, the DApp can securely use the token systems of existing blockchain
systems and execute functions efficiently with the component on manageable blockchain systems
(e.g., private blockchain) at the same time. Moreover, this process can be performed by extending
the existing smart contract code, such as Solidity in Ethereum. However, to support the new
program프로젝트로 돌아가기