It's got some pretty big names behind it, e.g. Creative Commons, Internet Archive, Wikipedia, Drupal, so the open source, all volunteer, Ourmedia initiative certainly merits some serious attention. The Internet Archive's founder, Brewster Kahle, has apparently pledged to preserve any material uploaded so I guess that suggests long term viability. If you want free, and apparently permanent, storage and bandwidth for your text, rich media (videos, audio, photos), and software then this open registry/digital repository may be of some interest to you. The only caveat is that you must be prepared to share your masterpieces with a global audience. When you upload material to Ourmedia you need to select from a range of licensing options, i.e. Creative Commons, Creative Commons Sampling, GNU Public License, Mozilla Public License, Netscape Public License, Public Domain or even a traditional copyright option (albeit that your material will be posted on a public web site designed for distribution).
Now I like the Ourmedia concept a lot, but the reality of working in higher education institutions is that persuading the majority of faculty to opt for this type of totally open sharing and reuse is pretty much an uphill task. If, however, there was support for groups and communities of practice then that would be a different matter. For example, in the medical arena there's a lot of material that's of high educational value but which would be perceived as unsuited to the type of totally open distribution offered by Ourmedia. Whether this argument is valid or not, matters less than that the belief will inhibit such useful and reusable material finding its way into the likes of Ourmedia.
I haven't thought this through fully as yet but it seems to me that there's a need for 'open' repository systems which build the confidence and experience of potential depositors of material by allowing them a greater degree of control over the permissions they attach to their deposits, e.g. public, private, friends, groups, communities. I suspect that, because the potential depositor felt more in control, in the long term the quantity of quality 'public' deposits would rise as the depositors moved through a continuum of 'private/my content' to 'public/our content' thinking. There are already hosted blog solutions and image archives which adopt this model; it's what turns the prosaic tool or service into something really powerful.
Another concern that would undoubtedly be raised is that, despite the involvement of some major players with a great track record, if Ourmedia disappeared tomorrow then those who had come to depend on it's presence could be left high and dry. The counter argument of course is that lots of copies keeps stuff safe and that the more the dissemination the greater the preservation.
There are of course plenty of top-down, one-way, repositories, e.g. The British Universities Newsreel Database (BUND) as well as other sources available via the British Universities Film and Video Council site. However, whilst these official collections are an invaluable part of our cultural heritage, what Ourmedia is tapping into is the belief that the artefacts produced by 'the others' has as much value to some people as does the official polished productions of the professional media machine. Alternatively, instead of a repository model there's the Blinkx TV approach which focuses on the search side of things, but with a twist. Blinkx claims their system 'watches' TV clips from participating sources and automatically builds contextual information and so makes it possible for a user to jump to a specific point in a remote clip.
But a reliable and remotely accessible repository in which to store and share rich media resources should be a pretty powerful part of the e-learning armoury. There are other examples of such user-supported bottom-up repositories. There's the Open Video Project but they seem to have stopped taking new submissions 'due to limited staffing' (see earlier comments about such risks) and I suppose we could also consider a general learning object repository like MERLOT as part of this genre as well. What the Open Video Project 'staffing problems' perhaps suggests is that the only sustainable way is to rely less on core staffing and more on making the users responsible for maintaining their own resources or collections thereof. I see the same challenges facing the podcast community, but at least here the directory node editor model of, say, an iPodder.org, makes for a more distributed and, therefore, perhaps more sustainable approach, i.e. no one individual or group is responsible for maintaining the whole system, but merely their part of it.
Despite the caveats above I applaud the Ourmedia inititative.