Re: [Gnu-arch-users] Re: Online book for usability

[James Blackwell]

>     > David Allouche Wrote:
>     > Essentially, that paper says: "You may know, as an uber-geek, that your
>     > tool is right and that new users get it wrong because the expect the
>     > wrong thing. Still, what matters is your users, so you'd better consider
>     > seriously whether doing the wrong thing user expects would not actually
>     > be the right thing." But it says so is much more numerous words in order
>     > to be convincing.

Tom Lord Wrote:
> Ick.  It's a confused sentiment.
>
> User interfaces are, in effect, "handles" on a particular device.
>
> Handles always _suggest_ to users how to use the device.  I recommend
> the classic "The Design of Everyday Things" by Donald Norman for
> principles like "natural mapping" (between control (UI) and
> functionality (interfaced device)).  If (like me) you have a stove on
> which you're constantly turning on the wrong burner by mistake, that
> book will shed new light on your life.

There's another way to look at things. If you run a hardware store that
only sells screwdrivers, and your customers keep bringing in broken
screwdrivers because they're misusing them as hammers, then you should
be selling hammers too.

WHen there is a disconnect between a tool and the user, the user or the
tool is broke (for simplicity, ignoring the edge case where _both_ are
broken). The book does a pretty good job of sidestepping the whole "is
it the user thats broken, or the tool" by pointing out that if enough
users are frustrated, the tool is essentially broken regardless of the
facts. 

I don't think the book quite gets to the point that it captures some
sort of universal truth. However, I think the logic is strong enough
that its worth having read and considered its arguments. 

> User's of hammers never (or rarely) make the mistake of picking up the
> wrong end and trying to hammer a nail with what was supposed to be the
> handle.  But if the head of a hammer were differently shaped -- looked
> and felt, accidentally, more like the handle than the actual handle,
> then that kind of mistake would be natural and people would make it
> all the time.  Even experienced hammer users would sometimes make that
> mistake.

Your hammer isn't a good example. A better example would be if some
crappy screwdriver company had taught all of society to use screwdrivers
for their hammering. Companies that made proper hammers would go out of
business (because everybody is using screwdrivers) and companies that
made proper screwdrivers (with handles that shattered when you hammered
with them) would also go out of the business. The market expectation is
that when it comes hammering, one uses a screwdriver with a very solid
handle.

> So when a device has the right functionality but users consistently
> try to use it the wrong way, it's the shape of the handle, not the
> underlying functionality, that is wrong.   The answer isn't to modify
> the device to do the wrong thing -- but to experiment with and tweak
> the shape of the handle.

Tom, the argument in the book is that if the tool is *already right*,
and the users are *already wrong*, then in order to compete effectively
one _must_ make the wrong tool.

Thats one of the strengths of the book. Book does a pretty good job of
sidestepping the good engineering vs. bad engineering, and shows that 
"even if you're right, you're still wrong" 

>     > Most people want to _do_ something which is not essential to the tool
>     > they are using. When the tool goes in their way and try to force them to
>     > change the way they work, their are frustrated, and may even become
>     > aggressive.
>
> Yes, ill-fitting tools cause stress among people who have to use them.

That's _exactly_ what the book hits on. That when it comes to user
expectations, using the proper design is the *wrong* thing to do,
because the product then no longer meets user expectations. 

Tom, I really do suggest you read the thing before commenting too much
on why it's right or wrong. Its not very long; I think its probably
around 30 pages in length.  I've been calling it a book because of its
format; not because of its size. 

I'm not promising that its right or wrong. I'm promising that it
provides a useful tool on one way to look at the problem.

-- 
James Blackwell          Try something fun: For the next 24 hours, give
Smile more!              each person you meet a compliment!

GnuPG (ID 06357400) AAE4 8C76 58DA 5902 761D  247A 8A55 DA73 0635 7400