[jira] [Commented] (GEODE-3710) Make CodecAlreadyRegisteredForTypeException a RuntimeException

["ASF subversion and git services (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/GEODE-3710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16184438#comment-16184438 ] 

ASF subversion and git services commented on GEODE-3710:
--------------------------------------------------------

Commit fb44ed928a7dd68e7c20db31f6f458a2ee1aa693 in geode's branch refs/heads/develop from [~gosullivan]
[ https://gitbox.apache.org/repos/asf?p=geode.git;h=fb44ed9 ]

GEODE-3710 Make CodecAlreadyRegisteredForTypeException a RuntimeException.

If we throw this exception, the `SerializationCodecRegistry` will
already be in an invalid state, so it doesn't make sense for anyone to
catch it.

* privatize `SerializationCodecRegistry.register()`.
* Throw a `GemFireConfigException` when a
  `CodecAlreadyRegisteredForTypeException` is thrown, because it is a
  fatal error that will halt client messaging.
* remove most of `CodecRegistryJUnitTest`.
  This test was mostly testing that `PowerMockito` could create enum
  variants that didn't exist.

Signed-off-by: Sarge <md...@pivotal.io>

This closes #825


> Make CodecAlreadyRegisteredForTypeException a RuntimeException
> --------------------------------------------------------------
>
>                 Key: GEODE-3710
>                 URL: https://issues.apache.org/jira/browse/GEODE-3710
>             Project: Geode
>          Issue Type: Improvement
>          Components: client/server
>            Reporter: Galen O'Sullivan
>             Fix For: 1.3.0
>
>
> We should make CodecAlreadyRegisteredForTypeException a RuntimeException.
> If we throw this exception, the `SerializationCodecRegistry` will already be in an invalid state, so it doesn't make sense for anyone to catch it.



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