on 16/12/00 00:35, Jeff Patton at jeff621@...
> Just to add my 2 cents to testability and UI - and I'm
> not sure it pertains directly to extremeModeling: I
> work in an XP shop, and indeed there is a focus on
> making things testable. I've seen it impose a
> detrimental effect on usability at times.
> Specifically, I've seen UI be implemented a particular
> way *because it's testable* not because it best
> supports what the user is trying to accomplish.
My comment wasn't about usability... It was about XP and lot's of people out
there saying that they are doing XP when in fact they found XP to be a
"cool" name to call their ad-hoc chaotic ways of working... [I'm of course
not talking about anybody in particular].
XP means taking every programming practice to an extreme. Let's say
testing... You have to test everything and you can't write code unless you
write a test case and try to break it. XP is also against any kind of
modeling at higher levels of abstraction (i.e. you can draw a couple of UML
diagrams if they represent implementation level models). This is highly
against any model-based approach to software development - including usage
centered design and others.
> Basically that points to keeping our eye on the goal -
> to create usable, good software. XP is a means to an
> end - not an end in and of itself.
If you mean that one can use some of XP practices within a different
development approach... That's Ok but that's not XP. The only thing new to
XP is taking every practice to an extreme... If you only take some you're
not doing XP... But you can develop good and usable software.
> --- John Sisk <john.sisk@...> wrote:
>> Hi Nuno,
>> You bring up an issue I hadn't considered fully , ie
>> testing the model.
>> Thank you for the link to www.extremmodeling.org
>> which opens up a new
>> learning path for me, and I'm sure some interesting
>> discussions here at work
>> John Sisk
>> Hi, the real issue regarding XP and any up-front
>> modeling approach is the
>> testability of the models. Can you build test cases
>> for analysis and design
>> level models? Whether they are for user interface
>> design purposes or not
>> doesn't matter.
>> Even if you argue that prototypes and subsequent
>> usability evaluation of
>> prototypes are a way of testing user interface
>> models... I don't believe
>> they match the initial idea behind XP... For
>> instance, can you break those
> Do You Yahoo!?
> Yahoo! Shopping - Thousands of Stores. Millions of Products.
Nuno Jardim Nunes
University of Madeira - Teaching Assistant
Mathematics Dep. - Computer Science Unit
phone: +351 291 705160 (direct) 705150 (secretary)
fax: +351 291 705199
Address: Campus Universitário da Penteada
9000 - Funchal - Portugal