RE: A simple guy with a simple problem

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Yes.  We can only make it work for 
the average idiot and the common-sense 
engineer.  The extraordinary 
idiot or the relentless expert beats 
us almost every time.

What makes web spec work so difficult 
is because there isn't one customer, 
one market, or one set of requirements. 
One dares to do less because it is 
actually really *dangerous* to do more. 
The freaking thing is an amplifier 
and if you turn on the power with the 
pots wide open, the feedback tears 
the speakers out of the stacks. 
On the other hand, when you need 
a big system and a lot of watts, 
you should know to turn it up 
one pot at a time and keep a hand 
on the Big Switch.   We can't do 
specs like that.  There is no Big Switch.

Our problem is this is the first time we've had an 
amplifier this big.  All the concepts 
of experience from previous work  
informs our designs, but the scale 
of the dammed thing defies any 
prior experience. We can experiment, 
but to what degree does the experimenter  
get access to the microphone?  

Compete or negotiate?  Compete and negotiate?
 
Who knew the remains of the Soviet 
army would turn to blackmail using 
the WWW once sensitive very valuable 
information was placed on it?  Some things defy our 
best efforts to guess and they are 
usually things we never thought 
would be an issue.

Oddly, it was the inability to secure 
the Internet completely and just 
that potential from the east that 
scared the bejeebers out of my CALS 
cohorts ten years ago.   When they saw 
the WWW design, they began to 
drop over dead like mine canaries. 
 
Their dieing word was:  competence.

So, back to schemas and the infoset.  
If schemas do less, will they do 
enough for everyone?  I don't think 
it is realistic to answer that here.  
Given any situation, a bad practice or 
a bad hair day will change the answer.

But best guess?

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Tim Bray [mailto:tbray@textuality.com]
Sent: Wednesday, March 14, 2001 3:04 PM
To: Bullard, Claude L (Len); Sean McGrath; xml-dev@lists.xml.org
Subject: RE: A simple guy with a simple problem


At 01:42 PM 14/03/01 -0600, Bullard, Claude L (Len) wrote:
>So to ask as straight man:
>
>Why does XML have a feature that permits 
>a "bad practice"?

Any tool powerful enough to be useful for professionals
is going to be powerful enough to be dangerous when used
improperly.  E.g. there are lots of examples of bad
practice that can produce poor results with chainsaws, 
procedural programming languages, and 18-wheeler trucks.
The alternative - not having them - is unacceptable.

<nostalgia>Anyone remember the person worrying out loud
here about the "billion laughs" self-exploding
XML instance?</nostalgia> -Tim