Re: dropped columns and ms access
От | David P. Lurie |
---|---|
Тема | Re: dropped columns and ms access |
Дата | |
Msg-id | cbl0sv$i4l$1@sea.gmane.org обсуждение исходный текст |
Ответ на | Re: dropped columns and ms access ("Philippe Lang" <philippe.lang@attiksystem.ch>) |
Список | pgsql-odbc |
Still having the same problems with 7.03.02.09 , but the deleted columns may turn out to be unrelated. I can load Access, and successfully run forms designed with linked tables, or open the tables in datasheet mode without error. The Access mdb containing the linked tables can be left loaded, I can come back after a period of time, try and open the linked tables in question and get the previous errors. Relinking, deleting and recreating links, or closing and reopening the Access mdb doesn't clear the error. Closing and reloading Access clears the errors each and every time. postmaster.log shows some errors that seem to appear once for each failed attempt to link to a table: ERROR: relation "msysconf" does not exist (This can be ignored) ERROR: invalid input syntax for type bigint: "" ERROR: invalid input syntax for type bigint: "" ERROR: invalid input syntax for type bigint: "" ERROR: invalid input syntax for type bigint: "" ERROR: invalid input syntax for type bigint: "" The space between the double quotes contains an up arrow when viewed from a cygwin bash prompt, and a tiny square when viewed in MS Word. Most of the tables have an int8 field that set up to function like an Access autonumber or SQLServer ident field, using a sequence, with default value of nextval('sequence_name'). Access doesn't have a native int8 data type, although Access projects can use int8 (bigint for SQLServer) data types with SQLServer tables. The ODBC driver is set to use row versioning, and all timestamp fields are timestamp(0) per recommendations for Access. The errors occur whether int8 is configured for default or numeric. I'm going to change the int8 columns to int4 columns and see if the problem still occurs. David P. Lurie
В списке pgsql-odbc по дате отправления: