[
https://issues.apache.org/jira/browse/CALCITE-2877?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16780089#comment-16780089
]
Andrei Sereda commented on CALCITE-2877:
----------------------------------------
[~julianhyde] this is in response to our discussion on [dev
list|https://lists.apache.org/thread.html/8c255842f5c8885d772c27546d5fe12dd4e2f2d124ac6ad221e47519@%3Cdev.calcite.apache.org%3E].
If using JNDI one has to make client code depend on JNDI plus we have to remove
connection from directory service when it is closed (to avoid memory leaks).
It seems to create more problems than solving.
{quote}
There’s nothing stopping you from creating schemas (or trees of schemas)
manually. And yes, you can put whatever kinds of objects you like as
constructor parameters for your schema. But if you want to use SchemaFactory,
or model JSON, or you want your plan to be instantiable in a different context
or VM, you have to play by the rules.
{quote}
> GeodeSchema change to make the constructor public to pass the client cahe
> directly
> ----------------------------------------------------------------------------------
>
> Key: CALCITE-2877
> URL: https://issues.apache.org/jira/browse/CALCITE-2877
> Project: Calcite
> Issue Type: Improvement
> Components: geode
> Reporter: Sandeep Chada
> Priority: Major
> Labels: pull-request-available
> Time Spent: 20m
> Remaining Estimate: 0h
>
> GeodeSchema change to make the constructor public to pass the client cahe
> directly
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
