We are currently developing a web-based application in Java for management of
digital object collections, and are building it on top of services exposed by
Fedora. We are using the SOAP interfaces for all Fedora communications.
We have had some questions about accessing datastreams that are reminiscent of
past discussions on fedora-users:
* we have realized that being able to assign an ID to datastreams would be
helpful for using the getDissemination method of API-A; it would allow for using
the getItem() method of the default disseminator (currently, fedora-system:3) to
request a particular datastream. In some past discussions, I saw that this
functionality was scheduled--is that accurate, and if so, is there a target
* an alternative way to get a particular datastream without an dsID is to bind a
datastream to a particular behavior, and publish that dissemination (right?).
That way, we just use the getDissemination() method of API-A to request the
datastream without having to send in a dsIDs since it has been bound ahead of
time. Is this possible / correct (deranged)?
* can anyone offer pointers on what it takes to write some behavior (say, as a
Java servlet) and declare and intergrate the bDef and bMech objects in Fedora.
I'm very curious how easy or difficult it is to start declaring and writing your
own behavior as Fedora objects, but have found almost no documentation on this.
Thanks in advance for any assistance!
ACS Technology Center