The organizational coalescence of Dspace and Fedora, announced yesterday, is welcome and not too surprising. This arrangement concentrates the resources of two organizations with similar missions and should strengthen their joint ability to secure future resources and to improve their products and services. The new organization is to be called Duraspace. The press release suggests that Fedora and Dspace software products will continue to exist independently, although presumably one might expect to see greater sharing of approaches between them.
The press release highlighted DuraCloud ...
Currently, the organization is developing DuraCloud, a new web-based service that will enable content replication to a network of cloud-based storage services, and provide an environment for running access and preservation services within the cloud. [DuraSpace Faq]
The idea here appears to be to provide convenient and mediated replication services for Dspace and Fedora users (and maybe others?). Other services might also be part of the mix and one can imagine a variety of those (format validation/migration, automatic classification, etc). (See the comment by Michele Kimpton from earlier this year ).
The continued trajectory of this organization will be interesting to watch. The transition from research-led, sponsor- and grant-funded activity to demand-led, revenue-generating activity is not straightforward, especially if one wants to preserve something of the ethos and impulses of the originating organisations.
It will also be interesting to see what actual revenue-generation potential DuraCloud has. The offer potentially simplifies the use of various cloud based services by hiding the multiple negotiation and other transaction costs behind a single service. How valuable is such a service to institutions? Given the early stages of academic cloud services, DuraCloud may have to create awareness about the model and its potential advantages as part of building the business.
A major question, presumably, will be to what extent there is sufficient volume of particular classes of data to make offering the service, and value-added services, 'routinizable'. Where data requires custom attention, or where the costs of providing particular services on it cannot be shared across a large number of customers, then the rationale for the service diminishes.
Anyway, congratulations to the leadership of the two organizations for taking this step and good luck with this new venture.
 Note: Confusingly, DuraSpace seems to have started out as the name for what became DuraCloud.