RE: [business-transaction] Email votes - 7 Issues - Ends Tues April 9

[Peter Furniss <peter.furniss@choreology.com>]

Bill

Your vote on

> > - Issue 107: Reply to CONFIRM_TRANSACTION if all is ok
>
> > NO     It seems counterintuitive to have a successful
> > CONFIRM_TRANSACTION get a message conditioned on whether
> > "report-hazard" is true.  A success is not a HAZARD.  If we want to
> > change the name of report-hazard to
> >
> "be-really-really-wordy-and-tell-me-everything-I-might-possibly-wa
> nt-to-know"
> > then this solution makes sense.  :-)

The thought was that it depends on whether the terminator is in the loop in
sorting out a muddle (rather than hazards being dealt with separately, by
management action, human intervention etc).  Setting "report-hazard" to
false means (and I think it says this somewhere) that the terminator is
asking to be informed of the *decision*, not of its eventual application.
The original application logic may not be able to cope with the multifarious
possibilities of hazard, in which case it's reasonable that it be able to
say it doesn't want to be bothered with being told.

(I'm not sure how happy I would be with an ATM screen that came up with a
message:
   "Something went wrong with the attempt to debit your account. It may or
may not have been debitted, or may have been debitted repeatedly. We are not
giving you any cash though."

I'd rather it said "transaction failed", and told the bank systems to sort
things out, involving humans if necessary.

Peter