"Am Thu, 31 Jan 2002 07:15:14 +0000 (UTC), schrieb
helebor@... ("Helen Borrie (TeamIBO)")

>If this isn't your issue, then another place to look will be to test (in your code, or via the debugger) whether the first dataset really did go into Edit mode. If you establish that it did, in fact, then test whether the second dataset succeeded in going into Edit mode.

Many thanks Helen for this hint and sorry Jason that I thought it was
caused by a change of IBO.
It was my fault.
I overwrote the edit method in my successor of the ib_query. There I
did something within a try..except and did also the inherited edit
within the try except without raising the exception to the calling
procedure.
Then in my application I do the edit within a try except and check the
errorcode in the except case and give the user the information that
the record is locked, if the errornumber is that of a locked record.
As I catched the exception within the edit method in my ib_query
successor, the edit in the application doesn't raise any exception and
so there is no errormessage displayed to the user.

Again, sorry for bothering you with this problem and thanks a lot for
giving me the decisive hint.

cu
Helmut