The testing drivers 8.4.0101 psqlodbc30a worked fine. Wonderful.
Is this a rollback to the previous behavior, or does this version use different code for the client side cursor? Or, I suppose this is a better question: Is this a permanent fix that will be included in the next official release, or is it a temporary fix just for me?
Thanks,
Tom
Hiroshi Inoue wrote:
Tom Goodman wrote:
Hi,
I just loaded the 08.04 psqlodbc drivers on a new machine, but ran into a problem where this version is improperly marking ALL fields in recordsets as read-only. I presume that this is related to the recent change that restored the "read only clause for read only cursors".
The final line in the procedure below fails with an ADO error message: (80040e21) "Multiple-step operation generated errors. Check each status value."
When comparing psqlodbc 08.04 and 08.03 I see that in 08.03 all of the fields of the recordset have the adFldUpdatable flag set in each field's .Attributes. In 08.04 none of the fields have this flag set.
Since this code runs without error when connecting to non-postgresql databases, I believe that psqlodbc is not properly setting the adFldUpdatable for the values of CursorLocation, CursorType, and LockType.
Is this a bug? Can it be addressed?
Could please try the drivers on testing for 8.4.0101 at
http://www.geocities.jp/inocchichichi/psqlodbc/index.html
?
regards,
Hiroshi Inoue