Re: John Galt: Lover of Standards

[Jill Burgchardt <jburgcha -at- pestilence -dot- itc -dot- nrcs -dot- usda -dot- gov>]

I think there are valid points on either side of this argument.

"A tale of two applications" 

Application 1 is an Informix database on a UNIX platform.  Updates to the 
software are released annually.  We've developed strong standards for the help 
and manual.  It has a reputation as a really solid, well-thought-out 
product--both the application and the documentation.  Without audience analysis 
and detailed business analysis, the system would be difficult to use.  
Relationships between screens are evident only with an understanding of the 
business.  Each release has refined and added to the business detail as 
requested by users.  The developers and organization are very proud of this 
application.

Application 2 is the complete opposite of Application 1. It is a web-based 
application delivered across multiple platforms/browsers.  The development team 
had approximately three months to get the initial release up and running.  
Updates are released weekly (more often if necessary). The help system follows 
"standards" (an html stylesheet) that was thrown together in an afternoon (just 
to provide a bit of consistency). Audience analysis was nonexistent.  (In some 
cases, so were the requirements.)  User feedback has been entirely 
"post-release."  That's not a problem, because of the frequent updates and 
web-based delivery.  If we had not delivered the initial product on time, or if 
the users had had to accept the initial release for a year, we'd be out of jobs. 
Instead, the product/team has a reputation for rapid delivery and responsiveness 
to change requests.  Again, the developers and organization are very proud of 
this application.

How much time/effort is it appropriate to invest in standards?  The classic 
tech-writer response applies:  "It depends."--on budget, timeline, what's 
necessary for safe use of a product, how intuitive a product is, frequency of 
releases, etc.

I can look at these two radically different products developed within the same 
organization and say that each was developed with an appropriate use of 
standards.  What would have been bad is deciding they both HAD to be done the 
same way because of a brain-dead policy that didn't recognize project 
differences. 


Jill Burgchardt
jburgchardt -at- pestilence -dot- itc -dot- nrcs -dot- usda -dot- gov