I fully agree with Mark's suggestion. A/V as well as Medical Imaging is not something you would like to deal with as a generic repository service. This does not mean that one doen't like to manage A/V and Medical images as part of an institutional collection in a repository. I believe we will see a much more service and network oriented approach. At the Erasmus Medical Center I am working with we are together with our National Broadcast Institute (Beeld en Geluid) managing video's for education. They have the complete video archive and services (like conversions, encoding) in place and can keep video data alive for indefinite periods.They lack (and that's where the repository challenge I believe lies) the specific metadata and search facilities we need for our researchers. This part is managed by our repository and the medical library. We can only manage the total service by working closely together.

In a similar way we are setting up a Research Archive based on DSpace where the images are managed by a medical image storage server. DSpace only manages the medical metadata (and the checksums of the image bitstreams). I am convinced that repositories will be moving in this direction. The major challenge and I believe this is what DSpace 2.0 will tackle is to manage the data over different (storage cloud) providers
and guarantee its integrity both from a technical/bitstream perspective as well as a format/information/relationship/meaning perspective. Some people don't get that this will not be one system/appication. It's a combination of services (including AV experts, eg. people) that will very likley not be working in your institution. In case of A/V this would likely mean storing 'raw' video bitstreams in a deep archive managed by an expert provider with scale advantages like BeeldenGeluid. They will under direction of the institutional repository transcode to any 'daily' format you like/need (flash, mpeg4 or whatever the future will bring) and take full advantage of the free hosting of these versions on Youtube and other providers. The 'only' thing the repository needs to manage is where which bitstream is and what metadata is linked to it in other systems/services....

A nice development but even more a major institutional organization challenge!

Peter Walgemoed

Mark H. Wood wrote:
If an organization has people whose focus is A/V production, then a
little help from them could go a long way.  And they probably also
wince at some of the stuff that people bring to us, and might like the
opportunity to help spiff it up.

For those who don't have a media department or can't call on it, maybe
the best DSpace can do is some ancillary scripts that package the
operations required to render such bitstreams in formats best suited
to wide use and long-term preservation.  I don't think we want it to
be automatic.  Some of us have a desire to archive both (say) the raw
MPEG-2 data for the sake of never losing anything, and a properly edited,
well-engineered, scaled and compressed version for daily use.

  

_______________________________________________ Dspace-general mailing list [email protected] http://mailman.mit.edu/mailman/listinfo/dspace-general
begin:vcard
fn:Peter Walgemoed
n:Walgemoed;Peter
org:Carelliance
adr:;;Boschdijk 250F;Eindhoven;;5612HJ;Netherlands
email;internet:[EMAIL PROTECTED]
title:Founder
tel;work:+31 40 2445257
tel;cell:+31 6 21206821
x-mozilla-html:TRUE
url:www.carelliance.com
version:2.1
end:vcard

_______________________________________________
Dspace-general mailing list
[email protected]
http://mailman.mit.edu/mailman/listinfo/dspace-general

Reply via email to