Skip to content
This repository has been archived by the owner on Sep 14, 2022. It is now read-only.
debrecenics edited this page Feb 11, 2016 · 18 revisions

MONDO Collaboration Framework

General architecture for collaborative modelling

Our architecture (illustrated below) assumes that different stakeholders may collaborate with potentially conflicting business interests, therefore protecting their intellectual properties is a critical challenge.

Collaboration Scenario

Our architecture uses global access control policies, which are uniformly enforced for both online and offline collaborative scenarios. Since the models can be stored over heterogeneous storages for these scenarios (e.g. graph databases, triple stores, version control systems), global access control policies cannot directly rely upon those of these underlying storages. For instance, if a model fragment is persisted as a file maintained by a version control system, then we are unable to provide fine-grained access control over the individual model elements stored in the file. Model-level attribute-based access control is envisaged for that purpose, which provides rule-driven, fine grained tuning of permissions.

A secure collaborative modeling framework provides customized secure views for different collaborators, which manage read and write access to model elements. Secure views can grant or deny access to individual objects, links and attributes in accordance with the global access control policy and regardless of the device used for collaboration (desktop computer, tablet, etc). Furthermore, all caches will sufficiently be emptied after collaborative session expire. As a result, intellectual properties of different parties can be respected in collaborative sessions.

Architecture of MONDO Offline Collaboration Framework

The architecture of MONDO Collaboration Framework in the view of Security Access is depicted below. This represents a client-server structure: underlying version control system (VCS), such as SVN or Git, stores the models in a serialized form. VCSs can provide high-level access control policies to files, repositories, branches, etc. (VCS ACL). These policies are reused in the MONDO Collaboration Server. Moreover, the high-level policies are extended with model-level attribute-based access control (ABAC) which are stored in another version control system (MONDO Access Control Storage). Main concept of ABAC is that a set of access control rules are used to permit/deny for a certain model element read/write operation for a specific user/role.

Architecture

For offline collaboration, an Eclipse based Modeling Tool is required which uses the models stored in its workspace locally. MONDO Collaboration Client establishes the connection to the Server and provides the collaboration specific functions for models including the DSE Merge model merging problem.

Documentation of MONDO Offline Collaboration Framework

  1. DSE Merge
  2. Workspace Tracker
  3. Setting up the environment - Offline
  4. Setting up the development environment - Online
  5. Description of LENS
  6. [Access Control Language](https://github.com/FTSRG/mondo-collab-framework/wiki/Offline---AC Language)