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

[James Blackwell]

I think Turnbull wrote this:
>     > Certainly, some of the principles are, but those also tend to be the
>     > principles that Tom (primus inter pares) explains regularly, some of
>     > them daily.  On the contrary, one of the book's main themes was "ask
>     > any five random people, they don't even need to be users yet, and do
>     > what makes them happy", which contradicts Tom's regular statement that
>     > even if everyone who posts to to arch-users agrees on a UI issue,
>     > that's of less importance than the instinct of any _one_ of several
>     > experienced developer/users.
>

Now, we listen to Tom Lord:
> "contradicts" or "qualifies"?
>
> Satisfying 5 random users is close enough to exactly what I do as to
> make for nevermind.   But what kind of "random" selection of users are
> we talking about?   Not very random but not deterministic, either.

The main point of the book was that during development, one should lend
more credence to what users expect than to the 'right' decision. Its a
derivation of "the customer is always right" 

The author then takes a bit of time explaining how one figures out what
the customers want. In order to rebut an expected argument asking
customers is too expensive and complicated, one can get a usable result
from querying as few as five random people.


> For me, it's very simple:

] (Tom gives a detailed description of how he gathers sampling from end
] users and how he weights the samples)

Yeah. That's exactly what he suggests what should be done. 

What really struck me about his book was that he presents a persuasive
argument that, when it comes to engineering software, often doing the
"wrong" thing is ironically right. He does this by showing us that 
a tool that the user doesn't understand how to use is a useless tool.
Case in point: Though he doesn't reference gnome, its does a pretty good
job supporting why gnome is doing the right thing by chasing microsoft.
Gnome should chase microsoft not because microsoft is doing the right
thing, but because it's what users expect.

His main focus is how to write software for profit, so the argument
doesn't carry to arch 100%. However, I still think its a useful tool
when it comes to proving one way or another whether or not that approach
is right. 

I think that the argument can be defeated by
  1. untying project success with financial success. This is easily
     done in the free software world. 
     
  2. arguing that educating users can either reduce or eliminate the 
     dependance of the end-user's understanding. This one is a bit
     more difficult to prove in light of his argument because, as he 
     states, "nobody reads the documentation." I think he's actually
     right when it comes to the common case. Most users do avoid docs 
     whenever possible.

Proving either of these two points would strongly obviate the need 
to write software geared for the expectations and understandings of the
common enduser.

-- 
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