RE: Missing F's (used to be XML.COM: How I Learned to Love daBomb)

[Eric Bohlman <ebohlman@earthlink.net>]

8/17/01 1:22:33 PM, Tim Greiner <majormit@WORLDNET.ATT.NET> wrote:

>Actually it's an old linguistics trick....the F's that most people miss are
>soft F's (with a V sound).

Yep.  The reason the trick works is that most people don't consciously read at the level of 
individual letters; their conscious perception is at the level of whole words or phrases.  In fact, 
I'd conjecture that poor readers would actually do better on the task than good readers, because 
the good readers are dealing with bigger chunks of text; they're working at the "forest" level 
while the poorer readers are working at the "tree" level, and it's easier to see the leaves (which 
is what the task demands) from the latter perspective.

Over on comp.lang.perl.misc, I've been arguing that there's a certain class of "stupid mistakes" 
that are much more likely to be made by smart people than by dumb people; they involve momentary 
lapses of attention that cause one to miss certain details (Perl allows trailing commas on lists, 
and some Perl programmers assert that using them is a Good Thing because it reduces the chance of 
adding a new item to a list and forgetting to put a comma after the previous last item; such lists 
are often written one item to the line (line breaks aren't significant in anything other than 
quoted literals).  Others were arguing that somebody who forgot to put the comma after the previous 
last item didn't have enough understanding of Perl syntax to be a professional Perl programmer).  
IMHO, the smarter someone is, the more likely he/she is to have "habitized" certain cognitive 
tasks, and the more susceptible they are to attention-lapse errors on those tasks.

A good piece of reading is Chapter 9, "Intelligence, or problem-solving ability" of Gerald 
Weinberg's classic _The Psychology of Computer Programming_.  Weinberg argues that a good 
programmer should understand enough cognitive psychology to understand what sorts of "stupid 
mistakes" he's likely to be susceptible to, and take deliberate defensive measures against them.

What "silly errors" do people here find they're most likely to make when creating XML documents?  
What "silly errors" do people here find they're most likely to make when writing code to process 
XML documents?  What measures are you taking to reduce the likelihood of making such mistakes?  
What aspects of the XML Recommendation and related specs do you think encourage such errors?  What 
aspects of the specs do you think discourage such errors?  What "silly errors" do you see most 
often in other people's work with XML?