Content uploaded by Michele Marchesi
Author content
All content in this area was uploaded by Michele Marchesi on Jun 02, 2017
Content may be subject to copyright.
Abstract—The Blockchain technology is reshaping finance,
economics, and money to the extent that its disruptive power
is being compared to that of the Internet and the Web in their
early days. As a result, all the software development revolving
around the Blockchain technology is growing at a staggering rate.
In this paper, we acknowledge the need for software engineers to
devise specialized tools and techniques for blockchain-oriented
software development. From current challenges concerning the
definition of new professional roles, demanding testing activities,
and novel tools for software architecture, we take a step forward
by proposing new directions on the basis of a curated corpus of
blockchain-oriented software repositories, detected by exploiting
the information enclosed in the 2016 Moody’s Blockchain Report
and the market capitalization of cryptocurrencies. Ensuring
effective testing activities, enhancing collaboration in large teams,
and facilitating the development of smart contracts all appear
as key factors in the future of blockchain-oriented software
development.
Keywords-blockchain; software engineering; smart contracts;
cryptocurrencies;
I. INTRODUCTION
In the past years, a lot of attention has been paid to
the emerging concepts of Blockchain and Smart Contract.
Organizations such as banking and financial institutions, and
public and regulatory bodies, started to explicitly talk of the
importance of these new technologies. Some observers are
even talking of the dawn of a new era, declaring that “we
should think about the blockchain as another class of thing like
the Internet [...]”[1] and that the “wide adoption of blockchain
technology has the potential of reshaping the current financial
services technical infrastructure.”[2].
From an economic and financial perspective, the overall
capitalization of digital currencies is about 12 billions USD, as
of October 2016, 80% of which is the capitalization of Bitcoins
alone1. Venture capital investments in blockchain startups has
been steadily increasing, from $93.8 million in 2013, to $315
million in 2014, to $490 million in 2015. Some people are
even saying that Bitcoin and blockchain remind them of the
Internet circa 1993, when huge amounts of venture capital
started to flow into Internet startups, eventually leading to
the emergence of companies such as Cisco, Yahoo, Google,
Amazon and others.
Alongside these good news, there are also bad news,
however. Ever since digital currencies started to represent
a real monetary value, also hacks and attacks started. The
“exchanges”, Web sites allowing to store digital currencies
and to trade them against other currencies, legal or digital,
1https://coinmarketcap.com/
experimented big attacks. The biggest was the MtGox attack
that occurred at the beginning of 2014, leading to a declared
loss of $600 million. The latest was the Bitfinex attack of
August 2016, with a loss of $65 million. Another remarkable
exploit was that sustained by the DAO organization in June
2016, leading to a withdrawal of Ether digital currency funds
worth $50-60 million. The Ethereum community defended
themselves from this attack by performing a hard fork of the
Ethereum software that forcibly recovered the stolen Ethers
and gave them back to the original owners. Almost all of
these attacks can be attributed to poor software development
practices.
The feeling many software engineers have about such in-
creased interest in blockchain technologies and, in particular,
on the numerous software projects rapidly born and quickly
developed around the various blockchain implementations
is that of unruled and hurried software development. The
scenario is that of a sort of competition on a first-come-first-
served base which does not assure neither software quality, nor
that all the basic concepts of software engineering are taken
into account.
This paper aims at revealing the current issues and new
directions for blockchain-oriented software engineering, and
investigating the need for novel specialized software engineer-
ing practices for the blockchain sector. To this purpose, we:
1) identified the most relevant challenges for the state-of-
practice blockchain-oriented software engineering;
2) highlighted peculiarities of some of the most popular
blockchain-oriented software projects going on in the
world;
3) proposed new research directions for blockchain-
oriented software engineering, based on the results ob-
tained from the previous steps.
The paper is organized as follows. In section II we report
the most relevant related work; in section III we present the
current challenges for blockchain-oriented software engineer-
ing; in section IV we provide statistical data on open source
blockchain-oriented GitHub repositories, whose selection has
been guided by the 2016 Moody’s report on the blockchain
sector [3]; in section V we propose new research directions
on the basis of the dataset analysis; and finally, in section VI
we report the conclusion.
II. RELATED WORK
In [1], Swan discusses the applications of the Blockchain
starting from the concept of cryptocurrency to that of the more
advanced Smart Contract, in an attempt to demonstrate that
Blockchain-oriented Software Engineering:
Challenges and New Directions
Simone Porru, Andrea Pinna
University of Cagliari
Department of Electrical and Electronic Engineering
[simone.porru, a.pinna]@diee.unica.it
Michele Marchesi, Roberto Tonelli
University of Cagliari
Department of Informatics and Mathematics
marchesi@unica.it, roberto.tonelli@dsf.unica.it
the Blockchain can prove to be as disruptive a computing
paradigm as the Internet and the mobile. When compared to
the Blockchain, however, mobile development has raised less
pressing issues from the software engineering perspective. In
fact, the need for mobile-specific software engineering tech-
niques has grown only recently, precisely as mobile applica-
tions started to be employed for business-critical purposes and,
therefore, to require enhanced security and higher quality [4].
Conversely, business is the main reason behind the creation
of blockchain-oriented software. The demand for security in
blockchain applications is, therefore, even more pressing, and
thus that for specialized software engineering processes.
III. BLOCKCHAIN -ORIENTED SOFTWARE ENGINEERING :
CHALLENGES
We define as Blockchain-oriented Software (BOS) all soft-
ware working with an implementation of a Blockchain. A
Blockchain is a data structure characterized by the following
key elements:
•data redundancy (each node has a copy of the
Blockchain);
•check of transaction requirements before validation;
•recording of transactions in sequentially ordered blocks,
whose creation is ruled by a consensus algorithm;
•transactions based on public-key cryptography;
•possibly, a transaction scripting language.
Considering the distinctive marks of a Blockchain, software
engineers could benefit from the application of BOS-specific
software engineering practices. Such practices would consti-
tute the base of a Blockchain-oriented Software Engineering
(BOSE).
In this section, we identify the most relevant BOSE chal-
lenges, and the issues which originate from them. To this
purpose, for most challenges, we also provide excerpts from
the SWEBOK2[5] to properly frame the related issues.
New professional roles.”A recognized profession entails
specialized skill development and continuing professional edu-
cation”2. Due to the business-critical nature of the Blockchain,
finance and legal subjects have shown increasing interest
toward BOS. At the same time, bootcamps for Blockchain
developers are flourishing. The Blockchain sector will need
professional figures with a well-defined skills portfolio com-
prising finance, law, and technology expertise. An example of
a new role could be that of an intermediary between business-
focused contractors with low technology expertise and IT
professionals.
Security and reliability.”Software Security Guidelines span
every phase of the software development lifecycle” and ”Soft-
ware Reliability Engineered Testing is a testing method encom-
passing the whole development process”2. A Blockchain must
guarantee data integrity and uniqueness to ensure Blockchain-
based systems are trustworthy. Ensuring security and reliability
in BOS development might require specific methodologies
such as Cleanroom Software Engineering [6] or thorough
2SWEBOK 2004 version
software reviews. Furthermore, mathematically sound analysis
techniques could help enforcing reliability and security-related
properties in blockchain-oriented applications.
Testing techniques can also enhance system security and
reliability. IBM recently expressed the need for continuous
testing techniques to ensure blockchain software quality 3.
Testing techniques should be based on the nature of the
application2which, in the case of BOS, is that of security-
critical systems. In particular, there is a need for testing suites
for BOS. These suites should include:
•Smart Contract Testing (SCT), namely specific tests for
checking that smart contracts i) satisfy the contractors’
specifications, ii) comply with the laws of the legal
systems involved, and iii) do not include unfair contract
terms.
•Blockchain Transaction Testing (BTT), such as tests
against double spending and to ensure status integrity
(e.g. UTXO4).
Software architecture.Specific design notations, macroar-
chitecture patterns, or meta-models may be defined for BOS
development. To this purpose, software engineers should de-
fine criteria for selecting the most appropriate blockchain
implementation, evaluating the adoption of sidechain tech-
nology, or the implementation of an ad-hoc blockchain. For
example, Ethereum5has adopted a key-value store, which is
a very simplistic database. By adopting a higher level data
representation such as an Object Graph, it would be possible to
speed up many operations which would otherwise be expensive
using a key-value store [7].
Modeling languages.Blockchain-oriented systems may re-
quire specialized graphic models for representation. More
specifically, existing models might also be adapted to BOS.
UML diagrams might be modified or even created anew to
account for the BOS specificities. For example, diagrams
such as the Use Case Diagram, Activity Diagram, and State
Diagram could not effectively represent the BOS environment.
Metrics.BOSE may benefit from the introduction of spe-
cific metrics. To this purpose, it could be useful to refer to
the Goal/Question/Metric (GQM) method, that was originally
intended for establishing measurement activities, but it can
also be used to guide analysis and improvement of software
processes [5].
Due to the distributed nature of the Blockchain, specific
metrics are required to measure complexity, communication
capability, resource consumption (e.g. the so-called gas in the
Ethereum system), and overall performance of BOS systems.
IV. BLOCKCHAIN -ORIENTED SOFTWARE REPOSITORIES
In order to define new research directions for the BOSE
on the basis of the state-of-practice of blockchain-oriented
software, we conducted an exploratory study on a corpus
comprising 1184 GitHub software repositories, which were
3https://twitter.com/ibmsoftware/status/776605297037172736
4Unspent Transaction Output
5https://www.ethereum.org/
identified with the use of the Moody’s Blockchain Report [3]
and CoinMarketCap 6. We first identified the most relevant
projects and players in the blockchain sector; then, we col-
lected metadata on the corresponding BOS repositories. Infor-
mation from the corresponding issue tracking systems were
also considered.7. In the remainder of this section, we provide
details about the methodology used to build the BOS corpus,
and the preliminary results we obtained.
A. Building a Dataset of Blockchain-oriented Software
In this paper, we define a BOS project as a software project
which contributes to the realization of a blockchain project.
This definition includes both blockchain platforms, such as
Bitcoin and Ethereum, and general blockchain software [8].
To identify BOS repositories we start from the corre-
sponding blockchain projects. Moody’s Investor Services re-
cently identified more than 120 publicly announced blockchain
projects in an in-depth report of the blockchain sector [3]. The
projects list covers rated issuers across financial institutions,
nonfinancial corporates and the official sector, and can be
considered as a comprehensive list of blockchain projects
going on in the world. The Moody’s list is not a list of software
projects; nevertheless, BOS projects stem from the blockchain
projects on the list.
In addition to the Moody’s list, we searched for the soft-
ware associated to the currencies and assets with the highest
capitalization, as reported by CoinMarketCap. Since we took
the Moody’s list as a baseline, we did not include currencies
and assets with a lower capitalization than Stellar, the least
capitalized cryptocurrency in the Moody’s list for which we
found a related software repository. Being Stellar on the 17th
position at CoinMarketCap, we focused on the first 17 most
capitalized currencies and assets.
Finding the software corresponding to a project in the
Moody’s list is not straightforward. When the project name
is within a list entry (e.g., The Hyperledger Project), the
software can be easily found by searching for the specified
project name. When not specified, we searched for the in-
volved blockchain startups, which often choose to publish their
software on code-hosting platforms (e.g. GitHub, Bitbucket).
As for the currencies and assets found on CoinMarketCap, this
process was easier since each list entry is linked to the official
website.
We focused on freely accessible, open-source software
hosted on GitHub, a platform hosting the vast majority of the
detected blockchain-oriented software projects. We decided to
only consider software hosted on GitHub repositories because
GitHub provides homogeneous metadata, which, in turn, allow
us to compare projects on the basis of standard features. For
instance, it is possible to evaluate project popularity by relying
on the amount of stars given to a project by GitHub users, or
on the number of forks stemming from it.
6https://coinmarketcap.com/. We refer to the market capitalization of
September 23, 2016
7We focused on Github-hosted projects (see section IV-A), which come
with the integrated GitHub issue tracking system
B. Dataset Analysis
At the end of the selection process, we identified 52 GitHub
accounts, which comprise 1184 repositories. We extracted in-
formation on popularity (Stargazers), programming languages,
community involvement (Contributors, Open Issues, Watchers,
Forks), and age (time elapsed since creation).
To focus on the most relevant repositories, we only con-
sidered those that i) are base repositories (not a fork from
a previously existing repository), ii) had been updated in the
previous 30 days, and iii) were created more than 30 days
before (i.e. have been modified at least once since creation) 8.
By using these criteria, we retained 193 repositories out of the
initial 1184.
C. Preliminary Results
Figure 1 reports the most used programming languages
among the 193 retained repositories. JavaScript, Python, Go,
C++, and Ruby are the top 5 languages, with BOS JavaScript
repositories accounting for more than 30% of the total. It
is interesting to note the presence of Python and Go in the
podium, especially in comparison with the number of Java
repositories–not reaching 4% of the total.
Table I shows that among the top 10 most popular repos-
itories (i.e. those with the highest number of Stargazers)
ethereum/mist and coinbase/toshi were created less
than 1 year and roughly 2 years ago respectively. As expected,
Bitcoin is the most popular project, neatly distinguished as
for stargazers (9966) and contributors (396). Ethereum is also
very popular, with three associated repositories in the top 10;
in particular, the one written in Go is just behind the main
Bitcoin repository. The top 10 BOS repositories were created
around 4 years ago on average, and most of them have a
considerable number of open issues. The statistics about forks
are staggering, topping at 4266 for the main bitcoin repository,
followed by the Go repository from Ethereum with 695 forks.
V. BLOCKCHAIN -ORIENTED SOFTWARE ENGINEERING :
NEW RESEARCH DIRECTIONS
Based on the results from section IV, we hereby suggest
some new research directions for BOSE.
Testing. A recent study on over 50000 GitHub projects
[9] has proved that a bigger team size leads to a higher
number of test cases, whereas the number of test cases per
developer decreases with an increase in the team size. It would
be interesting to investigate whether the same can be said
about BOS, considering that the most popular repositories
have an unusually high number of contributors, even for
open-source projects. For instance, almost 400 GitHub users
are contributing to the bitcoin/bitcoin repository, as
reported in Table I.
Collaboration. The high number of voluntary contributors tes-
tifies to the attractiveness of BOS in the open source landscape.
A large base of voluntary contributing members has been
shown to be a pivotal success factor in OSS evolution [10]. To
8All data were retrieved on September 23, 2016
Fig. 1. Languages across 193 repositories
TABLE I
EXTRACTED STATISTICS ACROSS THE TOP 10 BOS R EPOSITORIES
GitHub Repository stargazers contributors open issues age (days) watchers forks first language
bitcoin/bitcoin 9966 396 547 2105 1211 4266 C++
ethereum/go-ethereum 2160 78 285 1002 367 695 Go
ledger/ledger 1813 108 14 3055 103 255 C++
digitalbazaar/forge 1584 41 137 2260 103 241 JavaScript
ripple/ripple-client 1244 51 21 1437 968 486 JavaScript
ethereum/mist 1168 35 198 471 210 299 JavaScript
dogecoin/dogecoin 1153 300 52 1022 149 505 C++
ripple/rippled 1144 53 118 1782 246 338 C++
coinbase/toshi 839 18 97 749 98 187 Ruby
ethereum/cpp-ethereum 723 89 212 1001 196 270 C++
achieve sustainable development and improve software quality,
specific practices to enhance the synergy between the system
and the community would be highly beneficial to BOSE [11].
Enhancement of testing and debugging for specific pro-
gramming languages. Figure 1 shows that a number of
programming languages such as Go, Python, and Ruby are
gaining increasing popularity among BOS projects. This arises
the need for enhanced testing and debugging suites, tailored
upon the most popular BOS languages. Indeed, Java testing
suites have undergone much more testing than Go.
In addition, as BOS projects work with the Blockchain,
which is distributed by definition, testing in isolation would re-
quire properly mocking objects capable of effectively simulate
the Blockchain.
Creation of software tools for smart contract languages.
The implementation of Smart Contract Development Environ-
ments (SCDEs)–the blockchain-oriented declination of IDEs–
might be pivotal for the building and diffusion of BOS
expertise. Such environments could streamline smart contract
creation through specialized languages (e.g. Solidity, a lan-
guage designed for writing contracts in Ethereum).
VI. CONCLUSION
Blockchain-oriented software engineering must respond to
many challenges and define new directions to allow effective
software development. In the present work, we highlighted
the most evident issues of state-of-art Blockchain-oriented
software development, by advocating the need for new profes-
sional roles, enhanced security and reliability, novel modeling
languages, and specialized metrics. In addition, we used the
2016 Moody’s Blockchain Report and the market capital-
ization of cryptocurrencies to build a dataset of blockchain-
oriented software repositories. After retaining 193 repositories
out of 1184, we extracted statistical information on popularity,
collaboration, repository age, and programming languages.
On the basis of the results of the analysis, we proposed
new directions for blockchain-oriented software engineering,
focusing on collaboration among large teams, testing activities,
and specialized tools for the creation of smart contracts.
REFERENCES
[1] M. Swan, Blockchain: Blueprint for a new economy. O’Reilly Media,
Inc., 2015.
[2] Unicredit, “Blockchain technology and applications from a financial
perspective,” 2016.
[3] N. Caes, R. Williams, E. H. Duggar, and M. R. Porta, “Robust, cost-
effective applications key to unlocking blockchain’s potential credit
benefits,” 2016.
[4] A. I. Wasserman, “Software engineering issues for mobile application
development,” in Proceedings of the FSE/SDP workshop on Future of
software engineering research. ACM, 2010, pp. 397–400.
[5] A. Abran, J. W. Moore, P. Bourque, R. Dupuis, and L. Tripp, “Swebok:
Guide to the software engineering body of knowledge 2004 version,”
IEEE Computer Society, Los Alamitos, California, 2004.
[6] H. D. Mills, M. Dyer, and R. C. Linger, “Cleanroom software engineer-
ing,” IEEE Software, vol. 4, no. 5, p. 19, 1987.
[7] D. Larimer. Introducing bitshares object graph. [Online]. Available:
https://goo.gl/TWWSif. Accessed on 2016-10-20.
[8] G. Hileman, “State of blockchain q1 2016,” 2016. [Online]. Available:
http://www.coindesk.com/state-of-blockchain-q1-2016/
[9] P. S. Kochhar, T. F. Bissyand´e, D. Lo, and L. Jiang, “Adoption of
software testing in open source projects–a preliminary study on 50,000
projects,” in Software Maintenance and Reengineering (CSMR), 2013
17th European Conference on. IEEE, 2013, pp. 353–356.
[10] K. Nakakoji, Y. Yamamoto, Y. Nishinaka, K. Kishida, and Y. Ye,
“Evolution patterns of open-source software systems and communities,”
in Proceedings of the international workshop on Principles of software
evolution. ACM, 2002, pp. 76–85.
[11] M. Aberdour, “Achieving quality in open-source software,” IEEE soft-
ware, vol. 24, no. 1, pp. 58–64, 2007.
