[RBASE-L] - RE: [6.5++]: trigger/procedure strangeness

[Sikora, Charles]

Title: Message

Dawn,

 

I reset my values for LAST_MAINT to null (with an UPDATE of course ;>) , and then issued the mass update again. Same issue = no message saying that LAST_MAINT changed, and no update performed.

-----Original Message-----
From: rbase-l@rbase.com [mailto:[EMAIL PROTECTED]
Sent: Wednesday, September 28, 2005 12:03 PM
To: RBASE-L Mailing List
Subject: [RBASE-L] - RE: [6.5++]: trigger/procedure strangeness

Charley,

 

Maybe because the values of LAST_MAINT are not changing?  You said you're issuing the update, but the values are already there, so there's really no change.  Maybe the procedure only runs when the update causes a change in a value??

 

Can you update the LAST_MAINT column to null first and then issue the update as you have outlined?  Not sure of your procedure and how a null value will effect things - or maybe you could use some bogus date rather than null .. like 01/01/50.  Either way, the update to null or the bogus date will show you if the procedure works with the mass update.

 

Dawn

---

From: Sikora, Charles [mailto:[EMAIL PROTECTED]
Sent: Wednesday, September 28, 2005 10:10 AM
To: RBASE-L Mailing List
Subject: [RBASE-L] - [6.5++]: trigger/procedure strangeness

[Running RbWin 6.5++ on WinXP.]

 

I have an update trigger/procedure setup that checks for a change to a column called LAST_MAINT in table EXTINGUISHER.

If there is a change to the column, the procedure will perform an UPDATE to the column called NEXT_MAINT.  This was setup this way as opposed to a computed column because there are lookups involved regarding time extinguisher type and the maintenance periods based on that extinguisher type.

 

What is odd is that if I edit a row individually using EDIT * FROM x, or EDIT USING formname, and the column named in the procedure has changed, it will perform my update perfectly. I have a message pop-up in the post-update procedure that tells me the update is being performed in fact.

 

However, if I try a mass update at the R>:

 

 UPDATE table SET LAST_MAINT = value from column in table2 where column in table2 IS NOT NULL, the post-update procedure never runs, and thus the change is not made to the NEXT_MAINT column! Both columns are datatype DATE.

 

NOTE: I am only performing this mass update to force the procedure to calculate NEXT_MAINT for all applicable rows. I had already entered the LAST_MAINT dates before I wrote the trigger/stored procedure.

 

Is this behavior exhibited by the trigger/stored procedure normal for the mass update?

Thanks,

Charley