home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.databases.sybase
- Path: sparky!uunet!tekbspa!tssgate!esh
- From: esh@prague.tss.com (Edward S. Hirgelt)
- Subject: NULLs revisited
- Message-ID: <ESH.92Dec23103158@prague.tss.com>
- Sender: news@tss.com (USENET Network News)
- Organization: Teknekron Software Systems, Inc.
- Distribution: comp
- Date: Wed, 23 Dec 1992 18:31:58 GMT
- Lines: 28
-
- In rticle <ESH.92Dec15103611@prague.tss.com>, I wrote
-
- Edward> You can't change the definition of the table to prohibit nulls once
- Edward> you've defined it to permit them. You can, however, do one of two
- Edward> things:
- Edward>
- Edward> 1. Create a rule for the particular column that forbids nulls.
-
- It was gently pointed out to me (thanks Robert and Lisa of Sybase) that
- I am living in the past.
-
- If the value of a column is NULL, it is not tested against its rule.
-
- I knew this at one time (having worked on duplicating the behavior in
- APT), but I forgot.
-
- In short, NULL values do not fire rules. Triggers are still fired of
- course and remain the definitive way to ensure an altered column is not
- null.
-
- Hope this clarifies the problem without muddying it even more. I feel
- somewhat obfuscatory today -- too close to that December holiday I guess.
-
- Ed
- --
- Ex vitio alterius |Ed Hirgelt |esh@tss.com
- sapiens emendat suum. |Teknekron Software Systems, Inc|
- (Publilius Syrus) |Palo Alto, Ca. |
-