Triggers in 7.1
Date: Mon, 29 Aug 1994 07:38:53 GMT
Message-ID: <1994Aug29.073853.6415_at_dde.dk>
Hi there
I have tried to trace a senario where some triggers are 'replaced', and a driving changing SQL-statement is then submitted, and the following statement is found in the trace:
select baseobject,o.owner#,t.type,update$,insert$,delete$,refnewname,
refoldname, whenclause,actionsize,enabled,t.rowid
from trigger$ t,obj$ o
where t.obj#=:1 and t.baseobject=o.obj#
As far as I can see, this statement suggests that the sequence in which the triggers (on a given table, action and type), are selected (and firered) are pure random.
Does anybody know if this is a fair statement?
Anyhow one flag of warning should be hoisted: It is nessesary to record together with a total export, if this facility has been utillized, or the corresponding import might fail, if the database is not prepared for this issue.
-- Martin Jensen, Dansk Data Elektronik A/S, Herlev, Denmark E-mail: mj_at_dde.DK --- Don't take life too seriously -- you'll never get out if it alive.Received on Mon Aug 29 1994 - 09:38:53 CEST