[Sigia-l] Usability Testing comments from Giga

Thomas Vander Wal list at vanderwal.net
Thu Mar 27 15:18:45 EST 2003


Listera wrote:

>"Thomas Vander Wal" wrote:
>
>  
>
>>>We had several threads on this issue here over the past two years, so I
>>>won't rehash it. But, if every decision an IA makes has to be tested, one
>>>inevitably has to question the value brought to the table.
>>>
>>>      
>>>
>>Cute, but no.  This would mean that software developers that test every
>>application, object, and occasionally every line of code would not have
>>value.
>>    
>>
>
>It doesn't sound like you've ever written or managed code for a living.
>Designing systems/architectures is not the same thing as writing executable
>code, which literally stops executing if you leave out so much as a
>character or a digit.
>  
>
Wrong assuption.  I do manage code for a living. Just becuase an 
application runs does not mean it runs properly.  Based on requirements 
and design specifications have specific test applied to validate 
functionality and efficiency.  With large applications what seems like 
minor inefficiencies in code and have critical degredation  of speed. 
 Testing is much more than bug testing, which is what you are discussing.  

To this degree testing an IAs hypothesis is very similar to true 
application testing, not the simple bug testing you seem to think is the 
world of coding.  Yes, an IAs recommendation may work, but is it optimal 
for the users?  Have we set bench marks to validate?  Hence, we test. 
 We are professionals.

All the best,
Thomas




More information about the Sigia-l mailing list