You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@streams.apache.org by "Steve Blackmon (JIRA)" <ji...@apache.org> on 2017/07/31 20:57:01 UTC

[jira] [Created] (STREAMS-520) Standardize how accounts of interest are enumerated to Providers

Steve Blackmon created STREAMS-520:
--------------------------------------

             Summary: Standardize how accounts of interest are enumerated to Providers
                 Key: STREAMS-520
                 URL: https://issues.apache.org/jira/browse/STREAMS-520
             Project: Streams
          Issue Type: Improvement
          Components: Providers
            Reporter: Steve Blackmon
            Assignee: Steve Blackmon


Providers that pull details on a finite set of accounts are not currently consistent.  Some accept an array of object internal to the module (gplus and youtube) while others accept an array of strings (twitter, facebook, instagram).

There's no reason or benefit to this situation.  It would be easier to build a set of integrated compatible pipelines if we harmonized these pipeline configuration beans, and there would be less low-value config mangling code in each module to understand, test, and maintain.

I suggest that all providers that research a finite set of accounts supplied in a configuration object, should accept (only) an activity streams collection of optionally typed activityobjects, with either the id or some other identifying criteria populated such that the provider can locate the appropriate upstream object.

Also, the standard use of ids should use the activity streams format of the id rather than the native format.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)