There are some principles for deploying a digital heritage collection which 25 years of IT experience and supporting technical concepts make self-evident. Most of these are documented on this site, such as the idea of using virtualisation to separate out the various services that make up a digital collection, and the separation of the long-term storage of digital artefacts from their interpretation and display. In our case, we use DSpace software along with its robust database (Postgresql) to ensure that our storage warehouse of data meets the requirement of a long term repository. DSpace can, of course, simply be browsed for content, but from time to time, it may be desirable to raise a project to curate and interpret information contained in DSpace. I have always thought the display and interpretation requirement is often a short-term one, in comparison. So to my mind, a content management system like WordPress or Drupal would be perfect for such interpretation, simply dipping into DSpace to select the conformation to display.
The Roy Rosenzweig Centre for History and New Media, though, has developed Omeka, specifically to bridge the gap between appropriately rigorous digital storage and displaying collections of material. Omeka is developed and made available under the General Public Licence, so, as Free and Open Source Software, it provides future certainty to users, and the fact that it is free of cost makes it easy to deploy. While Omeka can be used as the underlying repository as well, this may be best considered as appropriate for small collections. It is clear, though, that the combination of DSpace for long term storage and Omeka for interpretation and deployment is a powerful one, which I hope to be exploring more in the near future.
Omeka is as easy to deploy as the blogging platform, WordPress and similarly uses a standard Linux (operating system), Apache (web server), Mysql (database) and PHP (development language) server. Similar to WordPress.com services, it is possible to use Omeka's own servers at omeka.net to deploy your own site, but most DSpace users will be probably have their own capability to deploy Omeka,
For projects like the Assynt Community Digital Archive, where a decision has been taken to keep the main DSpace repository off the public internet, Omeka offers the possibility of a displaying an informational sub-set of data, particularly useful for public relations or marketing purposes, or to fulfil a possible requirement of funders. Omeka also simplifies links with Zotero via a COinS plugin that generates bibliographic data, making adding bibliographic data to Zotero a click away. We have already discussed the usefulness of Zotero on this site, so deploying an Omeka site which makes it easy for browsers to populate a local Zotero database makes sense.
Omeka has an "Exhibits" plugin that allows one to use information from the collections Omeka contains to add interpretation. It seems to me that this is a particularly useful part of Omeka, and provides helpful separation of data from presentation. The "Exhibits" plugin seems particular useful to adding narrative to DSpace artefacts, but there will inevitably be an intervening step of getting DSpace information into Omeka which may be a clumsy way of managing the data.
There are some limitations or omissions to Omeka from my perspective. One would be that it would be good to have an Open Document Format viewer, allowing the display of uploaded ODF-format files. A workaround is that Omeka can display Google Doc information, but handing over data to third-party commercial entities does not sit comfortably with me. Another is that Omeka may be seen in some circumstances to fulfil the functionality of a fully-fledged storage repository like DSpace, so positioning it in a digital heritage project may be quite tricky (but on further thought, maybe that's a strength of Omeka). From a technical perspective, it would also be good to have a degree of database independence, or at least to offer support for Postgresql as well as MySQL.
Omeka is clearly a strong addition to particular digital humanities project deployments.
ADDENDUM:- Patrick Murray-John, one of the developers of Omeka, has been in contact to point out that the latest version of Omeka has programming interfaces designed to make it easier to exchange data with systems like DSpace. This seems to show an understanding of their target audience, and as the development team clearly found this blog post, an active interest in their project, both of which are positive points for anyone contemplating an Omeka deployment.