[
https://issues.apache.org/jira/browse/QUARKS-50?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15232261#comment-15232261
]
Dale LaBossiere commented on QUARKS-50:
---------------------------------------
This request only makes sense for cases where by definition of the operation
the returned/output stream is "logically identical to" the input stream - peek
and fanout. In fact TStream.peek() says that it returns "this" -- I think that
pretty clearly leads one to expect / implies the console should end up
rendering the returned stream the same as "this".
Definitely not Split nor Filter - that seems obvious to me.
Definitely not plumbing related things IMO. If there's some case where the
user wants that behavior they can manually add/duplicate tags as you described.
Is there a compelling case where duplication always makes sense?
> stream tags need to be "duplicated" through metric and fanout oplets
> --------------------------------------------------------------------
>
> Key: QUARKS-50
> URL: https://issues.apache.org/jira/browse/QUARKS-50
> Project: Quarks
> Issue Type: Improvement
> Components: Console, Runtime
> Reporter: Dale LaBossiere
> Labels: newbie
> Attachments: Screen Shot 2016-03-21 at 12.46.01 PM.png
>
>
> In a vintage 21Mar console's "static flow" view, stream (tag) coloring isn't
> retained through metric/peek and fanout oplets. That makes it more difficult
> to identify streams. That's particularly true in conjunction with the desire
> to minimizing the visual impact of injected metric oplets. See the streams
> associated with oplets 12,16,15,21 in the attached image.
> The output streams of those oplets are by definition the same content as the
> input stream so the input tags should be duplicated on the oplets' output
> streams. Seems like a runtime issue not a console issue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
