[
https://issues.apache.org/jira/browse/DERBY-6793?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14313029#comment-14313029
]
Myrna van Lunteren commented on DERBY-6793:
-------------------------------------------
Kristian worked on DERBY-5489, and it was noted in the release note that this
change would possibly cause incompatibility for applications retrieving the
same blob twice.
It seems this is documented incompatibility resulting from a necessary fix (the
alternative was incorrect data retrieval). Perhaps Kristian can add more
details. Or perhaps you have suggestions for a better approach...
> Stream or LOG value cannot be retrieved more than once
> ------------------------------------------------------
>
> Key: DERBY-6793
> URL: https://issues.apache.org/jira/browse/DERBY-6793
> Project: Derby
> Issue Type: Bug
> Components: SQL
> Affects Versions: 10.11.1.1
> Environment: does not matter
> Reporter: Sergey Zolotaryov
> Attachments: derby-bug.tar.gz
>
>
> Since migrating from derby 10.7.1.1 to 10.11.1.1 really innocent code stopped
> working: retrieving resultset column data using column names if resultset
> contains blobs stopped working.
> I am attaching a maven project with a unit test which proves the issue. If in
> pom.xml you change the version to 10.7.1.1 the test will pass. We are using
> spring-jdbc' queryForMap() which traverses resultset metadata and queries
> attributes by their column labels. Used to work before.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
