Development Overview

Overview

Heurist is a flexible research data management service, developed at the University of Sydney and running continuously on a number of public and private shared services since 2005. The project is extremely active - Oct 2019: more than 5000 commits since 2014, in use by dozens of research projects around the world with whom the development team has regular interaction and who inform development priorities.

Heurist provides a straightforward browser-based interface for building and publishing complex linked databases on a web service without programming. Its primary target audience is Humanities researchers, who often have complex data structure needs and limited technical support, but is equally suited to many routine documentation and data management tasks. It is not suitable for databases with >1M records or data which is primarily composed of simple numeric tables.

Heurist hides the complexity of the underlying SQL (currently uses MySQL and MariaDB) through user-oriented concepts. For example, a multivalue relationship with multiple target tables, hierarchichal relationship types, date ranges, notes and supporting references is represented as a 'relationship marker' field, defined in one minute with a few simple choices on a popup.

Main features

Heurist allows incremental on-the-fly structure modification - essential for projects which evolve through their lifetime - and long-term sustainability through a single shared codebase (currently running more than 100 projects) and generation of self-documenting archive packages. It builds and saves complex multi-level facet searches and network navigation rules through simple onscreen choices, allowing filtering, analysis and publication of subsets as maps, timelines, reports and a variety of export formats (CSV, XML, JSON, GeoJSon, Gephi, custom). Last but not least it can build a CMS website stored integrally as records in the database, embedding widgets in the website to filter and visualise the database content as required.

Our use of gitHub

Nostra culpa: The level of activity around Heurist is not immediately apparent from the gitHub profile. Our backlog is maintained in a Heurist database and Google Docs. We do a new release to the production services roughly monthly, although this does not translate to gitHub releases, nor have we got around to setting up issues, wiki etc. on gitHub. Yes we should .... and we will ... We are happy to share backlog information with potential developers. Please see "dev" branch for current code status.