[XML-DEV Mailing List Archive Home] [By Thread] [By Date] [Recent Entries] [Reply To This Message]

Re: Fallacies of Validation, version #2

• To: xml-dev@l...
• Subject: Re: Fallacies of Validation, version #2
• From: Bryan Rasmussen <bry@i...>
• Date: Thu, 26 Aug 2004 21:18:18 +0200
• In-reply-to: <200408261412.i7QECKS10981@s...>
• References: <200408261412.i7QECKS10981@s...>
• User-agent: Internet Messaging Program (IMP) 3.2.2

-- 
Bryan Rasmussen

> 
> >From Michael Kay:
> > The strategy (validating the user's address) assumes that 
> > you know better than your customers what constitutes a 
> > valid address. Let's face it, you don't, and you never 
> > will. A much better strategy is to let them (the user) express 
> > their address in their own terms. After all, that's what they 
> > do in old-fashioned paper correspondence, and it seems 
> > to work quite well.
> 

I have to say this is wrong in numerous levels, however only wrong in the
context of certain types of applications and certain types of developers.

example:

 Let us assume that I have a governmental application for the country of X,
addresses in X are to be formatted in a very specific manner, I know the rules
of all addresses that are valid addresses in the country of X. once an address
that is a valid X address comes into any X application it can be shared freely
between all X applications (I'm speaking technical feasibility here) if my
application is supposed to be restricted to addresses that are only X addresses
then if an address is not a valid X address it follows that either they address
has been
 a. filled out incorrectly by mistake, the client knows the proper address
rules
for the country of X but they mistyped
 b. filled out incorrectly by mistake, the client does not know the proper
address rules for the country of X, perhaps the client is a recent immigrant or
a student from abroad
 c. filled out incorrectly by malice.
 d. filled out correctly but the client is not actually able to be a client for
this service
If I know, as I do, that the address is not valid then I can save it for later
analytical processes to determine which one of these errors it is, but not save
it to do some particular thing for the client (interesting thing which in most
cases will not be done)

The problem is that applications are written which have as their target the
earth with the limits appropriate to a subsection thereof, which of course was
one of the examples that prompted Michael's observation. 

It would be nice to formulate this with some pithy saying, but I can't come up
with one at the moment. 

• Follow-Ups:
  • Re: Fallacies of Validation, version #2
    • From: Jonathan Robie <jonathan.robie@d...>

• References:
  • Fallacies of Validation, version #2
    • From: "Roger L. Costello" <costello@m...>

• Prev by Date: RE: are native XML databases needed?
• Next by Date: RE: Fallacies of Validation, version #2
• Previous by thread: Re: Fallacies of Validation, version #2
• Next by thread: Re: Fallacies of Validation, version #2
• Index(es):
  • Date
  • Thread