[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12430701 ]
Tomohito Nakayama commented on DERBY-1610:
--
I'm thinking to solve each those problem 1) to 5).
(May be more problem will be found in regards with what I
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429998 ]
Tomohito Nakayama commented on DERBY-1610:
--
I read the test class, parameterMapping.java as next.
This test is executed for each types on which column of
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429368 ]
Knut Anders Hatlen commented on DERBY-1610:
---
There will be user impact even if only a few of the incompatible combinations
are changed to throw
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429374 ]
Tomohito Nakayama commented on DERBY-1610:
--
I see
I think opinion of Knut is proper.
I will rollback DERBY-1535.
Then, this task is regarded as
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429387 ]
Tomohito Nakayama commented on DERBY-1610:
--
I think we need to fix the range of modification.
My plan is making compatibility around setBlob(notNull/null)
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429390 ]
Tomohito Nakayama commented on DERBY-1610:
--
I meant setBinaryStream(notNull/null) also ...
Updating column typed as CHAR to value passed via
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429416 ]
Tomohito Nakayama commented on DERBY-1610:
--
By the way, is it really possible to judge the type checks on the client side ?
There are only less
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429424 ]
Knut Anders Hatlen commented on DERBY-1610:
---
Yes, I think you can decide the correct type on the client side.
ParameterMetaData (which can be obtained
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429425 ]
Tomohito Nakayama commented on DERBY-1610:
--
Regarding check on the client side, I found description of PRPSQLSTT in the
spec of drda.
It seems that
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429427 ]
Tomohito Nakayama commented on DERBY-1610:
--
Oh...
Our comments were conflicted :)
I see.
Anyway it seems that I need to work on current implementation of
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429256 ]
Knut Anders Hatlen commented on DERBY-1610:
---
Hi Tomohito,
The BatchUpdateException indicates that the error happens when executeBatch()
is invoked, not
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429263 ]
Tomohito Nakayama commented on DERBY-1610:
--
The mystery was the reason why test of setNull generats different error message
than not null method in the
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429265 ]
Tomohito Nakayama commented on DERBY-1610:
--
I think release note should be prepared for this issue,
including difference originally introduced in
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429278 ]
Knut Anders Hatlen commented on DERBY-1610:
---
With the DERBY-1610_2.diff patch, setBinaryStream(null), setBlob(null) and
setNull(Types.BLOB) throw
[
http://issues.apache.org/jira/browse/DERBY-1610?page=comments#action_12429286 ]
Tomohito Nakayama commented on DERBY-1610:
--
Technically, I think we can make the calcJdbcTypeForNullValue() *more
complete*.
However, I'm not sure we
15 matches
Mail list logo