=Paper=
{{Paper
|id=Vol-1242/paper10
|storemode=property
|title=MDEForge: an Extensible Web-Based Modeling Platform
|pdfUrl=https://ceur-ws.org/Vol-1242/paper10.pdf
|volume=Vol-1242
|dblpUrl=https://dblp.org/rec/conf/models/BascianiRRSIP14
}}
==MDEForge: an Extensible Web-Based Modeling Platform==
https://ceur-ws.org/Vol-1242/paper10.pdf
MDEForge: an extensible Web-based
modeling platform?
Francesco Basciani, Juri Di Rocco, Davide Di Ruscio, Amleto Di Salle, Ludovico
Iovino, and Alfonso Pierantonio
DISIM - University of L’Aquila, Italy
{name.lastname}@univaq.it
Abstract. Model-Driven Engineering (MDE) refers to the systematic use of mod-
els as first class entities throughout the software development life cycle. Over the
last few years, many MDE technologies have been conceived for developing do-
main specific modeling languages, and for supporting a wide range of model
management activities. However, existing modeling platforms neglect a number
of important features that if missed reduce the acceptance and the relevance of
MDE in industrial contexts, e.g., the possibility to search and reuse already de-
veloped modeling artifacts, and to adopt model management tools as a service.
In this paper we propose MDEForge a novel extensible Web-based modeling plat-
form specifically conceived to foster a community-based modeling repository,
which underpins the development, analysis and reuse of modeling artifacts. More-
over, it enables the adoption of model management tools as software-as-a-service
that can be remotely used without overwhelming the users with intricate and
error-prone installation and configuration procedures.
1 Introduction
Model-Driven Engineering (MDE) refers to the systematic use of models as first class
entities throughout the software development life cycle. Model-driven approaches shift
development focus from code expressed in third generation programming languages
to models expressed in domain-specific modeling languages [1]. MDE increases pro-
ductivity and reduces time to market by enabling the development of complex systems
using models defined with concepts that are much less bound to the underlying imple-
mentation technology and much closer to the problem domain.
Over the last few years, many MDE technologies have been conceived for devel-
oping domain specific modeling languages, and for supporting a wide range of model
management activities. The relevance of MDE is evidenced also by the increasing inter-
est in many scientific endeavours, active technology projects, and numerous industrial
projects ranging from direct applications of MDE concepts and tools, to those develop-
ing its foundations [2]. Even though existing MDE technologies provide practitioners
with facilities that can simplify and automate many steps of model-based development
processes, empirical studies show that some barriers still exist for the wider adoption of
MDE technologies [3]. Among the main issues that currently hamper a wide adoption
of MDE there are at least the following:
?
This research was supported by the EU through the Model-Based Social Learning for Public
Administrations (Learn Pad) FP7 project (619583).
– the support for discovery and reuse of existing modeling artefacts is very lim-
ited. As a result, similar transformations and other model management tools often
need to be developed from scratch, thus raising the upfront investment and com-
promising the productivity benefits of model-based processes. For instance, when
modelers identify a need for a domain-specific modeling language, it is quite com-
mon to implement it from scratch instead of reusing already developed languages
that might satisfy their requirements;
– modelling and model management tools are commonly distributed as software
packages that need to be downloaded and installed on client machines, and often
on top of complex software development IDEs (e.g. Eclipse).
In this paper we propose MDEForge, an extensible modeling framework specifi-
cally conceived to address the issues previously mentioned. In particular, MDEForge
consists of a set of core services that permit to store and manage typical modeling arte-
facts and tools. Atop of such services it is possible to develop extensions adding new
functionalities to the platform. All the services can be used by means of a Web access
and by a REST API that permits to adopt the available model management tools as
software-as-a-service.
The paper is structured as follows: Section 2 discusses the motivations of the paper
by considering already existing works. Section 3 presents the architecture of MDEForge
and makes and overview of its main components. Two extensions of the platform are
presented in Section 4. Section 5 concludes the paper and discusses some research
perspectives.
2 Background and Motivation
The artefacts and tools that are typically involved when applying MDE approaches are
those shown in Fig. 1: editors are used to create models that in turn are manipulated
to generate other models or even code. Model repositories are employed to enable the
re-use of already specified models. The whole ecosystem is developed according to
different kinds of relations (e.g., conformance) with corresponding metamodels.
Fig. 1. Main MDE artefacts and tools
Managed Main purpose Typical deployment scenario
Artefact
AMOR [5] Model Model versioning Desktop application
Bizycle [6] Model Integration of software Desktop application
components
CDO Model Storage Client-Server application
EMFStore [7] Model Model versioning Client-Server application
GME [8] Model Storage Client-Server application
ModelBus [9] Model Model versioning Client-Server application
Morse [10] Model Model versioning Software-as-a-service
ReMoDD [11] Any Documentation Web-based interaction
Table 1. Overview of existing MDE tools providing storage features
Even though existing modeling platforms (e.g., EMF [4]) provide developers and
users with the tools shown in Fig. 1 they neglect a number of important features that
if missed the acceptance and the relevance of MDE in industrial contexts might be
reduced. In particular, in this paper we focus on the limited support for the re-use of
already developed modeling artefacts1 and for enabling the adoption of modeling tools
as-a-service. To this end in this section we overview existing works that are related to
MDEForge concerning the re-use of modeling artefacts (Sect. 2.1), and the possibility
to use model management tools as a service (Sect. 2.2).
2.1 Reuse of modeling artefacts
In this section we discuss state-of-the-art approaches (see Tab. 1) for providing reposi-
tories of modeling artefacts, and outline outstanding research challenges for achieving a
comprehensive solution to the problem of properly managing the persistence of models
and the deployment and discovery of any kind of model management tools to enable
their reuse and refinement.
AMOR - Adaptable Model Versioning [5]: it is an attempt to leverage version control
systems in the area of MDE. AMOR supports model conflict detection, and focuses on
intelligent conflict resolution by providing techniques for the representation of conflict-
ing modifications as well as suggesting appropriate resolution strategies.
Bizycle [6]: it is a project aiming at supporting the automated integration of software
components by means of model-driven techniques and tools. Among the different com-
ponents of the project, a metadata repository is also provided in order to manage and
store all the artefacts required for and generated during integration processes, i.e, exter-
nal and internal documentation, models and metamodels, transformation rules, gener-
ated code, users and roles.
CDO2 : it is a pure Java model repository for EMF models and meta models. CDO
can also serve as a persistence and distribution framework for EMF-based application
systems. CDO supports different kinds of deployments such as embedded repositories,
offline clones and replicated clusters. However, the typical deployment scenario consists
1
Hereafter with the terms modeling artefacts we include also modeling tools.
2
http://www.eclipse.org/cdo/
of a server managing the persistence of the models by exploiting all kinds of database
backends (like major relational databases or NoSQL databases), and an EMF client
application.
EMFStore [7]: it is a software configuration management system tailored to the specific
requirements of versioning models. It is based on the Eclipse Modeling Framework and
it is an implementation of a generic operation-based version control system. EMFS-
tore implements, in the modeling domain, the typical operations implemented by SVN,
CVS, Git for text-based artefacts, i.e., change tracking, conflict detection, merging and
versioning. It consists of a server and a client component. The server runs standalone
and provides a repository for models including versioning, persistence and access con-
trol. The client component is usually integrated into an application and is responsible
for tracking changes on the model, and for committing, updating and merging.
GME - Generic Modeling Environment [8]: it is a set of tools supporting the creation of
domain specific modeling languages and code generation environments. A repository
layer is also provided to store the developed models. Currently, MS Repository (an
object oriented layer on top of MS SQL Server or MS Access) and a proprietary binary
file format are supported.
ModelBus [9]: it consists of a central bus-like communication infrastructure, a num-
ber of core services and a set of additional management tools. Depending on the us-
age scenario at hand, different development tools can be connected to the bus via tool
adapters. Once a tool has been successfully plugged in, its functionality immediately
becomes available to others as a service. Alternatively, it can make use of services al-
ready present on the ModelBus. Among the available services, ModelBus also includes
a built-in model repository, which is able to version models, supports the partial check-
out of models and coordinates the merging of model versions and model fragments;
Morse - Model-Aware Repository and Service Environment [10]: it is a service-based
environment for the storage and retrieval of models and model-instances at both design-
and run-time. Models, and model elements are identified by Universally Unique Iden-
tifiers (UUID) and stored and managed in the Morse repository. The Morse repository
provides versioning capabilities so that models can be manipulated at runtime and new
and old versions of the models can be maintained in parallel;
ReMoDD - Repository for Model-Driven Development [11]: it is a repository of arte-
facts aiming at improving MDE research and industrial productivity, and learning ex-
perience of MDE students. By means of a Drupal Web application, users can contribute
MDE case studies, examples of models, metamodels, model transformations, descrip-
tions of modeling practices and experience, and modeling exercises and problems that
can be used to develop classroom assignments and projects. Searching and browsing
facilities are enabled by a Web-based user interface that also provides community-
oriented features such as discussion groups and a forum.
According to Tab. 1 the majority of existing approaches focus on providing support
for the persistence of models. Only ReMoDD supports other kinds of modeling arte-
facts, like transformations, and metamodels. However, the main goal of ReMoDD is to
support learning activities by providing documentation for each stored artefact. Con-
sequently, ReMoDD cannot be used to programmatically retrieve artefacts from the
repository or more generally cannot be adopted as software-as-a-service to search and
reuse already existing modeling artefacts. Most of the discussed approaches require
local installation and configuration. Only ReMoDD and Morse do not require to be in-
stalled locally. In particular, the modeling artefacts stored in ReMoDD can be searched
and browsed through a Web-based application. Morse provides developers with the
possibility to use it as a service.
2.2 Model management tools as service
The motivation or our work is shared also in [12] where authors propose the Modeling
as a Service (MaaS) initiative as an approach to deploy and execute model-driven ser-
vices over the Internet. This initiative is aligned with SaaS principles, since consumers
do not manage the underlying cloud infrastructure and deal mostly with end-user sys-
tems. Interestingly, in [13] authors investigate the problem of transforming very large
models in the Cloud by addressing two phases: i) model storage, and ii) model trans-
formations execution in the Cloud. For both aspect authors identify a set of research
questions, and possible solutions.
Even though there are different attempts [14] and projects (e.g., the EU MONDO
project3 ) related to the adoption of Cloud infrastructures to apply MDE, the area is still
mostly unexplored. In line with the MaaS initiative we want to contribute to this re-
search area with an extensible modeling framework that enables the adoption of model
management and analysis tools as service. As discussed later in the paper the frame-
work is at its early stages and we have not addressed yet aspects like workload and
capacity management that are typical in Cloud computing, however the results we have
obtained so far are promising.
3 Overview of the MDEForge platform
In this section we present the MDEForge platform that has been conceived to overcome
the issues discussed in the previous section. In particular MDEForge aims at:
– providing a community-based modeling repository, which underpins the devolpe-
ment, analysis and reuse of any kinds of modeling artifacts not limited to only
models;
– supporting advanced mechanisms to query the repository and find the required
modeling artifacts;
– enabling the adoption of model management tools as software-as-a-service;
– being modular and extendible;
As shown in Fig. 2 the MDEForge platform consists of a number of services that
can be used by means of both a Web access and programmatic interfaces (API) that en-
able their adoption as software as a service. In particular, core services are provided to
enable the management of modeling artifacts, namely transformations, models, meta-
models, and editors. Atop of such core services, extensions can be developed to add new
functionalities. For instance, by exploiting the transformation and metamodel services
3
http://www.mondo-project.org/
Fig. 2. Architecture of MDEForge
it is possible to extend the platform by adding a new service to enable the automated
chaining of model transformations as discussed in Section 4.
We envision different kinds of users of the MDEForge platform and in our opinion
they can be at least the following:
– Developers of modeling artifacts: as previously said we envision a community of
users that might want to share their tools and enable their adoption and refinement
by other users. To this end the platform provides the means to add new modeling
artifacts to the MDEForge repository;
– Developers of MDEForge extensions: one of the requirements we identified when
we started the development of MDEForge is about the modularity and extensibility
of the platform. To this end we identified a set of core services that can be used
to add new functionalities by means of platform extensions. In this respect, expe-
rienced users might contribute by proposing new extensions to be included in the
platform;
– End-users: a Web application enables end-users to search and use (meta)models,
transformations, and editors available in the MDEForge repository. Experienced
users might use the REST API to exploit the functionalities provided by the plat-
form in a programmatic way. For instance, tool vendors might exploit the function-
alities provided by their tools by exploiting some of the transformations available
in the MDEForge repository.
In the remainder of this section we give some details about the MDEForge reposi-
tory (Section 3.1) and the available core services (Section 3.2).
3.1 The MDEForge Repository
The Repository component plays a key role in the MDEForge platform and it has been
developed in order to store artifacts according to the metamodel shown in Fig. 3. In
particular, the repository has been developed with the aim of managing any kinds of
modeling artifacts (see the metaclass Artifact in Fig. 3). Each artifact refers to the
corresponding type, e.g., model, transformation, metamodel, etc. The specification of
the relation between a given artifact and the corresponding type is done by means of
the Relation elements. In turn, each relation is typed by means of a corresponding
RelationType element. By means of such modeling constructs it is possible e.g., to
specify the conformsTo relation between a model m1 and the corresponding metamodel
MM1 as shown in Fig. 4. Similarly, it is possible to specify any kinds of modeling
Fig. 3. Fragment of the MDEForge Repository Metamodel
Fig. 4. Simple content of the Repository
elements together with their relations. For example, it is possible to represent also the
execution engine of a given model transformation stored in the repository.
It is important to remark that all the artifacts stored and managed by the MDEForge
platform are related to the users that created them. The system permits also to make
artifacts public, private, or limit their visibility to specific users.
3.2 The MDEForge Core
In the following the core service previously mentioned and shown Fig. 5 are described.
Model Service: it manages models in the repository, thus it permits users to upload,
download, delete models, and even search models conforming to a specific metamodel;
Metamodel Service: it provides the means to upload, download, delete metamodels, and
find metamodels by a specific Universal Resource Identified (URI);
Transformation Service: it manages the transformations in the repository, i.e., it permits
to upload, download, delete transformations. Moreover, it permits to remotely execute
Fig. 5. Overview of the MDEForge Core services
transformations, or even find transformations by specifying the source and/or target
metamodel(s);
Editor Service: it permits to upload, download, or delete editors from the repository. It
is important to remark that at this stage of the project we manage the JARs containing
the implementation of a given editor (e.g., developed by means of GMF [15]). Conse-
quently, the editor service permits users to upload the implementation code of a given
repository to enable other users to find, download, and in case install them locally. In
other words, at this stage of the project we do not refer to on-line and collaborative ed-
itors even though this represents an important extension of the platform that we intend
to investigate in the near future.
A prototypical implementation of the MDEForge has been developed4 and even
though it is still at an early stage, we managed to apply the platform to support two
interesting problems: chaining model transformations, and measuring metamodels as
discussed in the next section.
4 Examples of MDEForge extensions
In this section we present two examples of extensions we have developed atop of the
core services of the platform and that have been successfully applied to deal with the
problems of chaining of model transformations (Section 4.1) and to calculate metrics
of metamodels to support the understanding of their characteristics (Section 4.2).
4.1 Automated chaining of model transformations
In [16] we have addressed the problem of automatically composing model transfor-
mations. In particular, we have proposed an approach to automatically discover and
compose transformations: developers provide the system with the source models and
specify the target metamodel. Then, by relying on the MDEForge repository, all the
possible transformation chains are calculated. Importantly, in case of incompatible in-
termediate target and source metamodels, proper adapters are automatically generated
in order to chain also transformations that otherwise would be discarded by limiting the
reuse possibilities of available transformations.
Fig. 6. Model transformation chain example
Figure 6 shows an explanatory model transformation chain. In particular, T1 is a
model transformation that generates models conforming to the target metamodel MM2
from models conforming to MM1 . Additionally, T2 is a model transformation that gen-
erates models conforming to MM4 from models conforming to the source metamodel
MM3 . In general, if the input metamodel of T2 would be also the output metamodel of
4
www.mdeforge.org
Fig. 7. Transformation Chain extension
T1 , then these two transformations could be chained. However, under certain conditions,
two transformations can be chained even though the output metamodel of the first trans-
formation does not correspond to the input metamodel of the second transformation.
To support the techniques presented [16] we have developed the extension shown
in Fig. 7 consisting of the Transformation Chain service that makes use of the Trans-
formation and Metamodel core services discussed in the previous section. In particular,
users specify the model to be transformed, the target metamodel by means of the Spec-
ify transformation chain operation. Such an input is used to calculate the transformation
chains that subsequently can be executed.
4.2 Measuring metamodels
In [17] we have developed an approach to measure metamodels with the aim of under-
standing their typical characteristics by investigating the correlations of different met-
rics applied on a corpus of more than 450 metamodels. In particular, we proposed an
approach for a) measuring certain metamodeling aspects (e.g., abstraction, inheritance,
and composition) that modelers typically use; and b) for revealing what are the com-
mon characteristics in metamodeling that can increase the complexity of metamodels
hampering their adoption and evolution in modeling ecosystems.
Fig. 8. Metrics extension
To perform such analysis we have developed an extension of the MDEForge plat-
form to support the calculation of metrics as shown in Fig. 8. In [17] we discussed met-
rics calculated only on metamodels, however we are working also on the identification
of possible correlations among transformation and metamodel characteristics. Finally,
to enable the management of the calculated metrics, the added service permit to export
CSV files encoding the values of all the calculated metrics. Generating CSV files en-
ables the adoption of statistical tools like IBM SPSS, Microsoft Excel, and Libreoffice
Calc for subsequent analysis of the generated data.
5 Conclusion and future works
In this paper we presented MDEForge, an extensible modeling framework supporting
the creation of a community-based modeling repository, which underpins the develop-
ment, analysis and reuse of modeling artifacts. The platform consists of core service
that can be extended and all of them are remotely available as software as a service
thus users are not overwhelmed with intricate and error-prone installation and con-
figuration procedures. Two concrete extensions and applications of the platform have
been presented. As future work we intend to investigate issues that are typical in Cloud
computing, e.g., scalability of the platform, and workload management. Moreover, we
intend to implement further extensions for instance to support advanced queries on the
repository. To this end we intent to investigate the integration of tools that have been
recently proposed [18]. Moreover, we plan to extend the platform by adding services
enabling collaborative modeling activities.
References
1. Schmidt, D.C.: Guest NOOPeditor’s Introduction: Model-Driven Engineering. Computer 39
(2006) 25–31
2. Di Ruscio, D., Paige, R.F., Pierantonio, A.: Guest editorial to the special issue on Success
Stories in Model Driven Engineering. Science of Computer Progr. (2014)
3. Whittle, J., Hutchinson, J., Rouncefield, M., Burden, H., Heldal, R.: Industrial Adoption of
Model-Driven Engineering: Are the Tools Really the Problem? In: MODELS. Volume 8107
of LNCS. Springer Berlin Heidelberg (2013) 1–17
4. Budinsky, F., Steinberg, D., Merks, E., Ellersick, R., Grose, T.J.: Eclipse Modeling Frame-
work. Addison Wesley (2003)
5. Brosch, P., Langer, P., Seidl, M., Wimmer, M.: Towards End-user Adaptable Model Version-
ing: The By-Example Operation Recorder. In: Procs.of CVSM ’09, Washington, DC, USA,
IEEE Computer Society (2009) 55–60
6. Kutsche, R., Milanovic, N., Bauhoff, G., Baum, T., Cartsburg, M., Kumpe, D., Widiker, J.:
BIZYCLE: Model-based Interoperability Platform for Software and Data Integration. In:
Procs.of the MDTPI at ECMDA. (2008)
7. Koegel, M., Helming, J.: EMFStore: a model repository for EMF models. In: Software
Engineering, 2010 ACM/IEEE 32nd Int. Conf. on. Volume 2. (2010) 307–308
8. Ledeczi, A., Maroti, M., Bakay, A., Karsai, G., Garrett, J., Thomason, C., Nordstrom, G.,
Sprinkle, J., Volgyesi, P.: The Generic Modeling Environment. In: Workshop on Intelligent
Signal Processing. (2001)
9. Hein, C., Ritter, T., Wagner, M.: Model-driven tool integration with ModelBus. Workshop
Future Trends of Model-Driven (2009)
10. Holmes, T., Zdun, U., Dustdar, S.: Automating the Management and Versioning of Service
Models at Runtime to Support Service Monitoring. In: EDOC. (2012) 211–218
11. France, R., Bieman, J., Cheng, B.: Repository for Model Driven Development (ReMoDD).
In: Models in Software Engineering. Volume 4364 of LNCS. Springer Berlin Heidelberg
(2007) 311–317
12. Brunelière, H., Cabot, J., Jouault, F.: Combining Model-Driven Engineering and Cloud
Computing. In: MDA4ServiceCloud’10 Workshop co-located with ECMFA. (2010)
13. Clasen, C., Didonet Del Fabro, M., Tisi, M.: Transforming Very Large Models in the Cloud:
a Research Roadmap. In: First International Workshop on Model-Driven Engineering on and
for the Cloud, Copenhagen, Danemark, Springer (2012)
14. Paige, R., Cabot, J., Brambilla, M., Chechik, M., Mohagheghi, P.: Procs. of CloudMDE -
First Workshop on MDE for and in the Cloud. (2012)
15. Eclipse: Graphical Modeling Framework. http://www.eclipse.org/gmf/ (2014)
16. Basciani, F., Di Ruscio, D., Iovino, L., Pierantonio, A.: Automated Chaining of Model Trans-
formations with Incompatible Metamodels. In: Procs. MODELS 2014 Accepted. (2014)
17. Di Rocco, J., Di Ruscio, D., Iovino, L., Pierantonio, A.: Mining metrics for understanding
metamodel characteristics. In: MiSE 2014 - ICSE Workshop. (2014)
18. Szárnyas, G., Izsó, B., Ráth, I., Harmath, D., Bergmann, G., Varró, D.: IncQuery-D: A
Distributed Incremental Model Query Framework in the Cloud. In: Procs. MODELS 2014,
Valencia, Spain, Springer, Springer (2014) Accepted.