OK, I think I see what you are asking. From my recollection, and from glancing at the code a little bit, I believe I implemented intra-node custody transfer. This transfer is doable since the end result is simply changing the authorizedName on the entity (from one publisher to another). However, inter-node transfer is the piece that requires the replication API. Since the replication API was never completed, neither was the full functionality of custody transfer. That is why the various JIRA issues were created.
I looked at the test code and I don't see any tests for intra-node transfer. The process you detailed in your message is exactly how a transfer works, but as mentioned only in the case that the entities (and associated publishers) are on the same registry node. I think perhaps we realized we weren't going to be able to do the full custody transfer yet so tests weren't created until it was fully implemented. On Tue, Apr 23, 2013 at 6:18 AM, Alex O'Ree <spyhunte...@gmail.com> wrote: > Jeff, it was just a general inquiry. Assuming that you've tested it > in some fashion, was there any kind of sample code that you had that > could be shared? I think I understand that pattern of interaction. > Registry A,,B > A publisher on A wants to transfer the ownership of some entity to a > publisher on B > A gets a transfer token > A, the person then uses the token and the keys on B to transfer the > entities, transfer_entities > transfer_custody (which is just keys) > when the transfer is complete, some kind of changeRecordNewData record > is sent from A to B, then B to A > > What's not clear is how the actual transfer of information occurs. > There's a few references to the replication API, but that's it. The > spec is a bit short on details (or I just simply missed it). Is the > replication API a requisite for custody transfer? > > > > > On Mon, Apr 22, 2013 at 1:35 PM, Jeff Faath <jfa...@gmail.com> wrote: > > I recall that I implemented that API as well but it was long time ago. > Are > > there any specific questions you have? > > > > As for the publisher to publisher transfer, that wouldn't be in the spec > > since a publisher is a juddi-specific concept to assign ownership to the > > various UDDI entities. > > > > > > On Mon, Apr 22, 2013 at 11:17 AM, Kurt T Stam <kurt.s...@jboss.com> > wrote: > >> > >> On 4/17/13 6:21 PM, Alex O'Ree wrote: > >>> > >>> what's the state of the custody transfer service? > >>> > >>> The uddi spec says its meant for intranode and internode, but I didn't > >>> see anything about within the spec about publisher to publisher > >>> trasnfer on the same uddi same node. I thought Kurt said something > >>> along those lines that that was another use case of it. I took a quick > >>> look at the code and it looks looks like it's implemented, despite > >>> https://issues.apache.org/jira/browse/JUDDI-272 > >>> > >>> who was working on this last? > >> > >> Jeff worked on it. ASAIK he said he completed the implementation, but > >> never added tests or examples for it: > >> https://issues.apache.org/jira/browse/JUDDI-256 > >> > >> Hopefully Jeff can provide more details. > >> > >> --Kurt > > > > >
