[
https://issues.apache.org/jira/browse/IGNITE-6055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16492799#comment-16492799
]
Nikolay Izhikov commented on IGNITE-6055:
-----------------------------------------
[~vozerov]
Please, review PR.
I've created a project for a compatibility check [1].
It checks that old client (2.5.0) can connect to new server (2.5.0-SNAPSHOT my
branch).
Two questions regarding this PR:
1. H2 has a datatype - {{LONGVARCHAR}}. It replaced by {{VARCHAR}} in Ignite.
Should I check {{LONGVARCHAR}} in tests?
2. {{ClientConnectionContext}} [2] and {{TcpClientChannel}} [3] contains equal
versions set. Should we refactor and leave just one set?
[1] https://github.com/nizhikov/ignite-thin-client-example
[2]
https://github.com/apache/ignite/pull/3973/files#diff-0f755e11a5c8734c51b951c2529ae87cR50
[3]
https://github.com/apache/ignite/pull/3973/files#diff-6d49b7001fdbb4d1522bd2e1ede6e654R72
> SQL: Add String length constraint
> ---------------------------------
>
> Key: IGNITE-6055
> URL: https://issues.apache.org/jira/browse/IGNITE-6055
> Project: Ignite
> Issue Type: Task
> Components: sql
> Affects Versions: 2.1
> Reporter: Vladimir Ozerov
> Assignee: Nikolay Izhikov
> Priority: Major
> Labels: sql-engine
>
> We should support {{CHAR(X)}} and {{VARCHAR{X}} syntax. Currently, we ignore
> it. First, it affects semantics. E.g., one can insert a string with greater
> length into a cache/table without any problems. Second, it limits efficiency
> of our default configuration. E.g., index inline cannot be applied to
> {{String}} data type as we cannot guess it's length.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
