[
https://issues.apache.org/jira/browse/HADOOP-6677?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12856234#action_12856234
]
Tom White commented on HADOOP-6677:
-----------------------------------
> This would be a pretty big change to the classification scheme that was
> argued out quite exhaustively in HADOOP-5073; there should be time for people
> to re-hash if they wish before anything is committed.
I agree that people should have a chance to discuss this change. As the
annotations begin to get more traction, it seems reasonable to be able to
change them to make them better fit how they are actually being used.
> If we're removing the limitation of who should use LimitedPrivate, and who
> can depend on it, then it's not really limited private anymore?
The code that is annotated LimitedPrivate gets to declare who may use it, so in
that sense it is still limited private, no?
> InterfaceAudience.LimitedPrivate should take a string not an enum
> -----------------------------------------------------------------
>
> Key: HADOOP-6677
> URL: https://issues.apache.org/jira/browse/HADOOP-6677
> Project: Hadoop Common
> Issue Type: Bug
> Affects Versions: 0.21.0
> Reporter: Alan Gates
> Assignee: Tom White
> Priority: Minor
> Attachments: HADOOP-6677.patch
>
>
> Trying to keep the list of all possible components sharing limited private
> interfaces up to date is painful. As other subprojects beyond HDFS and MR
> use this interface it will only get worse (see PIG-1311). If it is converted
> to a string other subprojects can use it without requiring patches to common.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
https://issues.apache.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira
