Assynt Digital Archive – System Change
|December 16, 2016||Posted by Stevan under Consultancy, Everyday Notes, Operations||
The Assynt Community Digital Archive was set up in 2011 as a long term community initiative. Readers of this site will know that among the founding principles of the Archive was an adherence to metadata standards and the choice of Free and Open Source software on which to run the Archive. Among the reasons for this, which have been proven over the years, is that Free Software finds no advantage in attempting to lock users in, the Free Software movement requiring technical excellence as the reason for staying with, or moving to or from, a software package.
The original choice of DSpace was an appropriate one for Assynt’s archive, as at that stage it was not clear how the archive would develop, or how the community would make use of it. DSpace would provide an industrial-strength, proven platform to hold securely any digital item thrown at it.
Another contender was the Omeka system, also discussed extensively on this site. At the time the Assynt Archive was set up, Omeka was still in its infancy, having been developed from around 2008. Like DSpace, it adheres to standards, most notably using Dublin Core metadata, as well as providing standard interfaces to export and import data. Unlike DSpace, which seeks only to be a repository of information, Omeka also allows extensions for creating digital exhibits, and provides mechanisms to view some details of the digital record easier. There are pros and cons regarding this philosophical difference in approach, which need not be discussed here.
One advantage Omeka has, in practical terms when it comes to managing a digital archive in the longer term, is simplicity of technical implementation. Omeka is based on PHP and MySQL (or MariaDB) while DSpace is based on Java and Postgresql, requiring a runtime Java environment and a more complicated build and upgrade routine. While it would be wonderful if Omeka abstracted the database requirement, allowing Postgresql to be used as an alternative to MySQL, there is no doubt that it is easier to run an Omeka instance than a DSpace instance.
Meanwhile, in Assynt, we are gearing up for some additional demands of the digital Archive, as the Coigach and Assynt Living Landscape Partnership (CALLP) has been successful in gaining lottery funding for a series of large projects over the next five years. It s likely that these projects’ data legacies will be met by the Assynt Community Digital Archive, and there was a desire to ensure that a suitable degree of flexibility, to support better the projects’ potential requirements, was available to them. It looked likely that Omeka, now a much more established system, would be a better option.
A series of meetings and demonstrations of the systems with the volunteer archivists gave the go-ahead to transfer one DSpace community to Omeka, to see whether it suited general requirements. That was successful, and all the DSpace communities and collections were transferred, using import plugins on the Omeka side, and standard export interfaces on the Dspace side.
Among the advantages are a slightly more acceptable, to casual browsers, user interface – Omeka looks more modern and less academic in its standard guise – and the fact that many data types, especially images, are displayed in line with the record. From a systems architecture point of view, it was decided to provide each existing (and future) DSpace community with its own Omeka instance, so that each community or project can make its own decisions about how look-and-feel, or even about where it wants to host its data. This approach does make for more repetitive future systems administration, but also spreads any future risk.
As before, all the new systems were built using virtual machines, making the systems less hardware-dependent. Hardware, in some circumstances for community archiving, is now much less of an issue compared with a few years ago, and Omeka’s lighter requirements contribute to this.
It looks as though cross-training to Omeka would be easy, and training new volunteers would be no more challenging than with DSpace.
The change to Omeka from DSpace for the Assynt Archive by no means implies than Omeka is “better” than DSpace, or that Dspace is not fit for this purpose. It simply means that, given this choice of option, for Assynt’s current requirements, Omeka looks to fit the bill a little better.
If you want further information about this, please contact me.