[Tfug] Speaking of desktops (little 'd')...

[Bexley Hall]

Hi, Paul,

--- On Tue, 11/11/08, Paul Lemmons <paul at lemmons.name> wrote:

> thread did. I think that there is an important point that is
> being missed here, though. It may have been addressed in the
> afore mentioned thread but I quit reading after a while.
> That point is that there is an ephemeral and elusive element
> required to create the perfect desktop, user interface, OS,
> whatever, that simply does not exist. That is a complete
> knowledge and complete respect of a user's preference.

Yes -- but only to a point.  E.g., the "user interface" in an 
airplane cockpit is highly controlled (regulated?).  Pilot A
might *wish* the radio controls were in a different place than
Pilot B -- but, it ain't gonna happen.  There is some value
in the fact that ALL pilots *know* where to find things regardless
of the aircraft they are flying.

[this seems an extreme example... i.e., you can't physically
*move* the radio each time a different pilot gets in the
aircraft.  But, its only a matter of time before all of these
controls become "virtual"]

> For example (Not to pick on Don, the example just happens
> to be handy):
> 
> "...highest on my [] List is "never steal the focus"..." 
> is a preference. It is not a natural law. There are no
> physics to say that this is the way it should be. It is the
> way Don thinks. Don is absolutely and unequivocally right,
> but only for Don and the people that agree with him.

Sorry, I should have been more explicit:
   What are the "inexcusable sins" encountered in current
   "user interfaces" (sorry for the imprecision but intend that
   to encompass all sorts of interactive devices).
The parenthetical remark is the key, here.

I consider stealing the focus to be inexcusable (hence my later
question:  "does it EVER make sense to steal the focus?") as
it presupposes that the user is watching the display for ANY
"window" that might pop up anyWHERE and steal the focus.

I design "things" (tangible objects -- embedded systems) so it is
not uncommon for there to be "eyes-free operation" (i.e., unlike
a PC where you are staring at the screen constantly).  So, for
example, in a process control system, the user may be watching
some mechanism and using cursor keys to adjust its position
(he most probably doesn't care that the carriage is displaced
3.902" to the left and 12.846" above it's 'origin'... he's just
trying to get the cutting tool positioned close to the 'work'
he's machining).  Allowing something to steal the focus is
counterproductive (in this case) *and* opens the door for errors

E.g., the user is prepared to hit some key -- ENTER? -- as
part of the operation he is engaged in... and, one ohnosecond
before doing so, a window steals the focus and that "keystroke
in process".  So, you have to then impose rules on any window
that could asynchronously assert itself:
- there can be no "default action"
- disable any input for some period of time so you KNOW the
  window is mapped *and* the user isn't in the middle of making
  an action (keystroke) intended for the window that *had* the
  focus

And, even in a PC environment, many people still watch the keyboard
(if they aren't touch-typists) *or* the "copy" they might be
transcribing (e.g., typing in a document; transcribing numbers
from paper invoices; etc.).  So, it seems pretty common for people
to *not* have their attention on the screen 100% of the time -- yet,
the focus can be stolen regardless of where their eyes are...

As I said, I just can't conceive of anything short of panic()
that could merit the use of this capability.  :<

> Personally it drove me crazy when I upgraded from GG to HH.
> In GG when I clicked on a link in T-Bird it would whisk me
> over to the browser on another screen to show me my results.

But I don't see that as stealing the focus.  You were clicking
on a link whose default action was assigned to: "view in browser".
You *expected* to end up *in* a browser -- either one that was
already open or one that would magically appear.  With your eyes 
closed, you could continue working *after* having clicked on that link
with no surprises (?)

> In HH it quit doing that. It would open the page in FF but I
> had click to it to see it. I am as happy as a clam in the
> sand that II has brought this feature back.

<frown>  As above, I just don't see this as "stealing the focus".
It's like setting the mouse to automatically position itself
on the default control in a dialog that pops up... if that dialog
can only pop up as a result of some action you have taken, then
this is just a shortcut.  You and it both *know* where your
attention is going.
 
> So, one of us, Don or I will most likely be unhappy because
> one of us is not going to get what we want. What is a
> programmer to do? We can offer configuration options. But
> wait, my wife hates having to configure every little thing.
> Ok, a theme? Maybe but there will be elements that the
> designer thought made perfect sense and I think are the
> dumbest thing since candy cigarettes. Programmers do try
> hard but it is not a battle that can actually be won.

Some devices (remember the snipette of my OP that I included above)
are really hard to configure.  Do you want to be able to reconfigure
your phone (iPhone) "user interface"?  Rearrange the keys on the
numeric keypad so '0' is at the top?

Other devices (like the aircraft) really *shouldn't* be configurable
as many people need to use them and *know* what to expect.  E.g.,
I designed a user interface for a process control system that was
"highly configurable".  You could bind "process variables" (e.g.,
"inlet air temperature", "pan rotation rate", etc.) to arbitrary
"indicators" (e.g., seven-segment displays, analog "meters",
pie charts, etc.).

It was *glorious*!  :>

Except, no one could use it -- because you (as an "operator") -- would
walk up to a machine that someone else had been running (i.e., had
previously "set up") and everything was in the "wrong" place.  In
the end, it was easier to just sell that configuration ability as
a "service" to customers:  "How would you like the user interfaces
on your machines configured?" and take that away from the user entirely.

> Approaching this from the other side, as a user, I can ask
> the same question. What is a user to do? The answer I give
> is find what you like and use it. My wife chose Windows. It
> would be a very poor choice for me. She likes it. It is
> right for her. I like kubuntu. I am even getting to where I
> like KDE4. I have one son that likes Vista and another that
> likes OSX. (I thought I raised them better ;-)). Every one

Hmmm... do any of these kids, by chance, bear a strong resemblance
to the MAILMAN??  :>  (apologies if that is too off-color)

> of these people are intelligent people who have made their
> choices based on what they like; not something that can be
> quantified and measured.

Hmmm... I'm not sure I agree there.  I can't believe that all
the user interfaces we interact with on a daily basis were just
"arbitrarily" designed.

The phone/calculator keypad always bugs me.  Someone (some *two*!)
obviously made a concious decision to layout the keys for one
in one order and the other in a different order.  Was the first
"arbitrary" and the second, "Oh, boy!  We sure goofed on that first
one!  Let's fix it for *this* one..."?  Were *both* arbitrary?
Or, were both *deliberate* with studies of their respective
uses?

Recall the early Sharp (?) wizard with its "alphabetical order"
keypad.  I suspect someone conciously decided NOT to use a
QWERTY layout:  "We're targeting casual users" (remember, this
predated the ubiquitous nature of the computer) "so we should
probably put the keys in an order that makes it easier for
people to find the letter they want to type..."

> I have used all of those systems myself. They all have
> strengths and weaknesses. Some I like more than others. None
> of them are "wrong". I am not sure there truly
> really *is* a "right" and "wrong". I do,

So, if there isn't a "wrong", then isn't any "reasoned approach"
a "right" (or, at least, a "notwrong")?

> however, think there is a "Like" and
> "Dislike" and no amount of math, science or
> in-depth studies, education, evangelism, brow-beating or
> name calling will ever change that.

Again, I don't know.  I have to believe there is *some* science
lurking behind all this -- instead of just "art".  Giving the
user infinite freedom is just abrogating your respnsibility as
a designer (and, pushing all that onto the user).

I'm *sure* there are some folks who wish the brake was on the 
right and accelerator on the left (why *are* they in their
relative positions, anyways?).  Hell, some folks want the
steering column on the "wrong" side of the car!  ;-)

> This turned out to be more of a rant than I had intended
> when I started it so I am going to stop. My apologies. All I
> really wanted to say is I like what I like and for the life
> of me I could not tell you why. I can say that most of my
> choices are not based on technological reasoning.

But, you admit there is a definite *feel* for what is "right",
right?  I.e., if someone sat you down in front of something
"different", it has the potential to "feel" wrong (it could
also end up feeling a *different* sort of "right")?  So, there
is something (intangible?) that you acknowledge as being present
in the interface that makes it work (e.g. Sharp's first keyboard
vs. their revised QWERTY) or not?