[
https://issues.apache.org/jira/browse/GEODE-7694?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17014712#comment-17014712
]
ASF subversion and git services commented on GEODE-7694:
--------------------------------------------------------
Commit 8ddc0e84745ec983f2554425ec3316f35951d2fa in geode-native's branch
refs/heads/develop from Blake Bender
[ https://gitbox.apache.org/repos/asf?p=geode-native.git;h=8ddc0e8 ]
GEODE-7694: fix pdx type lookup (#572)
- < operator should have used typeId, was using className which is NOT UNIQUE
- This should fix problems with __GEMFIRE_JSON PDX type
- fall back to comparing className when both typeIds are 0
- Local regions appear to operate with typeId == 0 for everything
- Local regions will still be broken for __GEMFIRE_JSON type, because className
is the same for different types
> PdxType comparison should favor typeId rather than class name
> -------------------------------------------------------------
>
> Key: GEODE-7694
> URL: https://issues.apache.org/jira/browse/GEODE-7694
> Project: Geode
> Issue Type: Improvement
> Components: native client
> Reporter: Blake Bender
> Priority: Major
> Time Spent: 40m
> Remaining Estimate: 0h
>
> The `operator<` method of PdxType in the native client simply compares the
> class names, which is incorrect and leads to the fact that, in NC, you cannot
> currently have two separate PDX types with the same class name. The
> comparison should favor type ID first, then if both type IDs are 0 (true for
> local regions), fall back to class name. This will still fail for local
> regions in the scenario where class names are the same, but it's much, much
> better than we have now.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
