Archives of the TeradataForum
Message Posted: Mon, 18 Dec 2006 @ 15:29:16 GMT
| Subj: | | Re: BEFORE Statement Trigger |
| |
| From: | | Victor Sokovin |
| | Whilst now an academic exercise, I still think this is an interesting question, and no doubt something I'll have to return to in the
future. | |
I think triggers as they are implemented now belong in the academic exercise arena and nowhere else. See, for example, one of the recent
threads:
www.teradataforum.com/teradata/20060420_092109.htm
To me it was interesting not only because of the bug described in the original posting (I wonder whether there is an update on it) but also
because of the limitations in the BEFORE trigger we found while trying to work around the bug.
The limitations are probably necessary as the previously implemented "flexible" triggers must have caused a lot of grief to transactional
processing and databases in general. It is not for nothing that ANSI are trying to be strict with them. But it will take time for database vendors
to re-implement and it will take time for us as users to re-educate ourselves.
Regards,
Victor
| 
