Issue 2672: Explicit "Link and Unlink Operations needed? (pids-rtf2) Source: (, ) Nature: Uncategorized Issue Severity: Summary: Summary: Issue: Does the spec need explicit "Link and Unlink Operations, or just guidance on how to assert the DuplicateIDs and ExternalIDs traits? Resolution: Revised Text: Actions taken: May 28, 1999: received issue Discussion: End of Annotations:===== Reply-To: "Jon Farmer" From: "Jon Farmer" To: "pids-rtf2" Subject: We can start now! Date: Fri, 28 May 1999 11:23:08 -0400 MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 Content-Type: text/plain; charset="iso-8859-1" X-UIDL: 3f0fe8ee008d570d7e743f2a003d6866 Hello all, Our mailing list was broken at the omg until today so that nothing could propagate out. Please understand therefore that the silence was not evidence of lack of zeal for this thing. Here is our timeline: 3/26/99 Establish RTF in Philadelphia 7/31/99 Comment Deadline 8/27/99 Report Deadline in San Jose Let's now generate the issues list! The rtf can only deal in changes that fix errors or add clarification. Therefore for each issue anyone of submits, we must be prepared to successfully cast it as such. For example, here is the first few issues I want to submit: Issue: Does the spec need explicit "Link and Unlink Operations, or just guidance on how to assert the DuplicateIDs and ExternalIDs traits? Since a given client (which in typical practice may be a program that scans and detects dupes) may be aware of two duplicate IDs - but NOT the duplicates the PIDS server is already aware of - if this client asserts duplicate IDs, is the semantics "these are all the dupes" or "these are two more"? HL7 has distinct operations for merge, unmerge, link, and unlink. Personally, I think that the argument by symmetry suggests we add link/unlink operations with exactly the same returns and arguments as the merge and unmerge. These ops would go on the IdMgr interface. I would cast it as a fix because as it stands I think th service is open to destructive update on th ExternalIDs and DuplicateIDs traits. By theway, I must confess that I have to go back and read the spec to figure out whether this discussion pertains only to one of the two traits.