SIGIA-L Mail Archives: Re: SIGIA-L: BDUF vs AD
Re: SIGIA-L: BDUF vs AD
From: George Olsen (george.olsen_at_pobox.com)
Date: Wed Aug 15 2001 - 13:58:44 EDT
At 12:35 PM -0400 8/15/01, scott paterson wrote:
>Tim, Can you also elaborate on what Big Design Up Front is. I have never
>heard this term before. Thanks for the indulgence.
Not sure how Tim is using the term, but BDUF is a term often used by
proponents of Extreme Programming (XP)
<http://www.cutter.com/ead/ead0002.html> and other "agile" software
development processes -- and they usually use it in a negative sense.
Essentially they're reacting against the traditional enterprise
software approach, where the entire spec is done upfront, the design
is locked down and the programmers are just there to execute the
spec. At best, they're advocating an iterative approach, at worse
it's a rationalization for the "code and fix" approach that many
programmers would rather do. (And certainly like any other movement,
the leaders often have a much more nuanced approach than what their
followers may advocate....)
But certainly the record of software development prompts some
re-examination of the process. Take this example from Chad
Dickerson's "CTO Connection" in InfoWorld
<http://iwsun4.infoworld.com/articles/op/xml/01/08/13/010813opconnection.xml>:
>WRITING SOFTWARE is very, very hard and I hate to
>disappoint everyone, but in my opinion, there is
>little hope that software development practices will
>take us much farther in the next 10 years than they
>have in the past 10. Each year since 1994, the
>Standish Group has released its Chaos Study, a sort of
>"state of the union" of corporate software development
>practices. In 1994, the first results indicated an
>industry in disarray: Only 16 percent of software
>development projects finished on time, within budget,
>and with the full functionality initially planned at
>the project's outset. The most recent results of the
>study for 2000 showed significant improvement for this
>metric -- all the way up to 28 percent.
>
>I can only imagine how my parents would have reacted if
>after scoring a 16 percent on an exam early in the
>school year, I had run home beaming with an
>improvement to 28 percent.
The proponents of "agile" outline their views in their manifesto at
<http://www.agilealliance.org/> I've only done some investigation
into XP, so I don't want to make definitive conclusions yet, however,
I reached similar conclusions as software engineering guru Ed Yourdon
<http://www.yourdon.com/articles/0107cw.html> that so far XP seems to
have worked well on small (up to a half-dozen developers, 3-6 months)
but there's questions about how well it will scale up to larger, more
complex projects. And I'm a bit dubious about XPs arguments that they
can entirely do away with specialists (including UI designers) in
favor of cross-trained programmers.
To me, like many movements that are reacting against something, the
"agile" proponents have probably swung too far over into the other
direction. Martin Fowler gives a more nuanced approach
<http://www.martinfowler.com/articles/designDead.html> blending the
best of both old and new -- but I'm afraid a lot of programmers in
the field that I've met often take the parts they like, and ignore
the parts they don't. For example, XP does stress listening to your
users, but often I'd heard this dismissed as BUFD. (Of course I'd
argue that a lot of the traditional requirements development that's
been done was flawed precisely because they _didn't_ do a good job of
listening....)
But there's definitely some good ideas in the "agile" development
movement. (There's actually an interesting parallel to some of the
conflicts between the traditional CHI usability people and people
who've come into the user experience design arena via the web. For
example, the CHI-Web list had a huge debate after Jared Spool
announced he's got some preliminary data suggesting that some of the
most usable sites on the web were created without traditional
usability specialists or "recognized" usability methods.)
Anyway I think it's good for IA/UIs to learn a bit more about the
issues, since you're likely to encounter programmers who are
advocating it.
--
________________________________________________________
George Olsen george @ interactionbydesign.com
User Experience Architect 310-993-0467
http://www.interactionbydesign.com
UX weblog & essays http://www.interactionbydesign.com/thoughts/
This archive was generated by hypermail 2.1.2
: Sun Nov 23 2003 - 22:54:47 EST
|
