[OASIS Issue Tracker] (OSLCCORE-46) Discovery of services through query rather than LDPC traversal

[OASIS Issues Tracker <workgroup_mailer@lists.oasis-open.org>]

    [ https://issues.oasis-open.org/browse/OSLCCORE-46?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=61545#comment-61545 ] 

Martin Pain commented on OSLCCORE-46:
-------------------------------------

Can those use cases be solved (inefficiently) by traversal?
If so, I see this as an optimisation, and therefore not suited to the core spec. (It's been suggested before that any optimisation that we try to put in won't end up helping most users, so it's better that we don't try and "micro-optimise" and end up complicating the spec.)
We can recommend that servers implement query capabilities for discovery resources, but I still feel that anything beyond that belongs in a profile or extension, probably vendor-defined (but preferably public), as they will know what actually optimises the requests for their tool. Then others can re-use that if it's public.

> Discovery of services through query rather than LDPC traversal
> --------------------------------------------------------------
>
>                 Key: OSLCCORE-46
>                 URL: https://issues.oasis-open.org/browse/OSLCCORE-46
>             Project: OASIS OSLC Lifecycle Integration Core (OSLC Core) TC
>          Issue Type: New Feature
>            Reporter: David Honey
>            Assignee: James Amsden
>
> The proposed 3.0 discovery requires that clients traverse a hierarchy of LPDCs to find service(s) of interest. Some OSLC 2.0 clients may [erroneously] assume they can discover such services through a single GET of the service provider catalog. When the OSLC client has significant network latency with the OSLC provider, having to iteratively GET LDPCs might lead to slow discovery performance. 



--
This message was sent by Atlassian JIRA
(v6.2.2#6258)