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

RE: Re: Are the publishing users happy? Why not?

• To: 'Mike Fitzgerald' <mike@w...>, xml-dev@l...
• Subject: RE: Re: Are the publishing users happy? Why not?
• From: "Bullard, Claude L (Len)" <clbullar@i...>
• Date: Fri, 21 Feb 2003 13:41:58 -0600

True enough but it is common.  I was just looking at 
a data specification that insists in Louisiana all 
first, last, middle names plus suffix can be stuffed 
into a 20 character wide field.   It ain't practical 
in a land where a lot of the names have French origins.
So the programmer has to call the State contact and 
ask "Is this REALLY what you want" meaning, "Do you 
understand that we will truncate and you have no way 
of knowing how much of or which part of the concatenated 
string is going away?".

And that is a simple one. (Of course, they shouldn't have 
concatenated anyway but that is a different problem.)

Still, I am a believer in cheap validation if one can 
get the rules to match the real world data.  Admittedly 
and by example, that can be tough to do in advance. A 
goodly part of the cost of records management implementation 
is finding all the bad rules over good data problems.

len

From: Mike Fitzgerald [mailto:mike@w...]

Michael Kay wrote:
> I've always been sceptical about validation. It encourages people to
> enter incorrect data in order to get it past incorrect rules. It's one
> of the things that gives computers a reputation for being inflexible.

Sounds to me like its more of a data model/web design problem than a
validation problem.

• Prev by Date: RE: Re: Are the publishing users happy? Why not?
• Next by Date: Re: The subsetting has begun
• Previous by thread: Re: Re: Are the data users happy? Why not?
• Next by thread: Google as Big Brother
• Index(es):
  • Date
  • Thread