[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-look
Subject:    Re: User Interface.. If you ever wrote or will write a software GUI, then READ THIS!
From:       Sergej Malinovski <sergej () nospam ! dk>
Date:       2001-03-10 11:57:55
[Download RAW message or body]

Christian A Str¸mmen [Number1/NumeroUno] wrote:
> I recently got hold of a page on the web that changed the way I look at
> user interfaces, this was a really extreme experience for me, and I really
> believe that this article (/book) is something EVERY KDE hacker should
> read.

Niet. No. I think you should read everything you can find, not just a
specific article like this one. Actually, I don't see this one as one of the
best written. It's not a book, as he would like it to be, it's a rant. And
just that.

It's mostly just blah blah, and it won't really help an average KDE developer
that much. He talks too much about (very specific Windows) problems and too
little about solutions. Understandable, we all can get carried away, but if
this article is supposed to be helpful, it's not. However, it will make you
think about the UI problems, if it's your first article on the subject...

Besides, the author confused Emacs with Vi ;-) Well, all I am trying to say
is don't believe everything you read (including this), as Dave allready
pointed out.

I have to admit that I disagree with many of the things this guys writes
bout, though he seems to have read all the latest books on UI design (from
his book section). I also don't like his attitude towards *NIX and the
"clueless" programmers.

Just to name a few things:

His first axiom is flawed (or simply not complete).
He confuses "easy to learn" with "easy to use".
He is oversimplifying too many things, like chapter 3 and 6.
He confuses variable-width fonts with small/bad fonts (and therefore suggests
a wrong solution).
His suggestion in chapter 8 about acting stupid to test things is simply
stupid.

> 1) About four years ago, many windows started sprouting three little ridges
> on the bottom right corner which look like a grip. It looks like the kind
> of thing somebody would put on a slide switch to increase the friction. It
> affords dragging. It just begs to be dragged to stretch the window. Look at
> picture resize-grip.gif

Well, the current problem with this in both GNOME and KDE is that resize
functionality is the job of window manager, but the resize widget is *in* the
window and should fit into the space of other widgets like status bar and
scroll bars. I like MacOS9's implementation of this more than Windows'. I
would like to hear how this is going to be solved...

> 2) If you read what's on the last part of page 4, you'll see that the
> settings dialog that appears in for example kmail and konqueror are really
> bad, tabs are much more preferable (please read this yourself, and you'll
> get the point)

Niet. Dave allready told you why. I will elaborate a little. The authors
thinking is a little flawed in that he uses simple examples.

The problem with the new iconstyle listing is that it that there is less
visible indication about what a click on a particular icon will do than with
tabs. Also when it's harder to see what's currently selected (again, only
compared to tabs).

> 3) The less text the better! Less less less less less!!! (page 6)

Sometimes.

> 4) A dropdown combobox should show ALL of the entries, it SHOULD NOT
> CONTAIN A scrollbar!!! (page 7)

The scroll bar in itself isn't bad. What's bad about it is, that if you miss
to hit it, you either select something you didn't want to, or you click
outside and everything disappears.

What is bad? The long list are bad. Too many choices to make... So let's try
not to make very long lists. Well, sometimes there is no the option but to
have a great many things in a list. So what can be done?

Now, what has already been done to solve this old problem? Auto-scrolling,
recent items at the top, scroll bars, manually resizable list... come to my
mind. Perhaps there are others.

I am against auto-scrolling. Much of auto, or more precisely much of
mouse-over is irritating for me. And auto-scrolling is irritating. Ever come
across those fancy web pages that use auto-scrolling instead of regular
scroll bars? Screw them! :) Auto-scrolling just doesn't give you enough
control to scroll trough all the mess -- this is it's biggest flaw as I see
it. You cannot control the scroll speed, you cannot jump from top to bottom
of the list etc. So perhaps in combination with regular scroll bar, it could
actually be useful. This needs to be tried out somehow... What do think?

What about recent items at the top? I like the idea, it works as intended and
is actually useful. Normally, even if there are many choices like in a long
list, you never use more than a few, at least not use constantly. The font
list i a classic example of this. I don't think it is necessary to
elaborate...

Scroll bars, see auto-scrolling (and what's bad above). They give the control
to fast choose the item you want. The problem could be scroll bars that are
too small. Well, this -- the size -- should be configured globally like in
Windows. To many users the default sizes are okay, but to others (with bad
sight), they are too small. KDE should make it possible to configure
(globally).

How about manually resizing the popped-up drop-down list? Internet Explorer
allows this for it's URI drop down list... I never seem to this feature. I
don't find it to be very useful, but if someone does, it should remember the
new size.

> 5) Use monospace fonts for editboxes!  Fonts with variable spaces are not
> good for this! (page 7)

As I said, I don't totally agree. I make fonts, so think the problem is font
size and quality and not the type. An old man in his example has as much
trouble controlling the mouse as reading that small text!

> Last, some wisdom words from the article (/book):
> "A user interface is well-designed when the program behaves exactly how the
> user thought it would."
> "Every time you provide an option, you're asking the user to make a
> decision. Asking the user to make a decision isn't in itself a bad thing.
> The problem comes when you ask them to make a choice that they don't care
> about." "Users Don't Read the Manual."
> "Users can't control the mouse very well."

I will "attack" the above statements another time...

From the article: "When Microsoft first usability tested the tabbed dialog
interface, usability went up from about 30% (the old Mac way) to 100%."

How do you measure usability in percent? Hehe, how about a new satiric KDE
banner "KDE 2.1 - now with KDevelop 2.1 and 2.1% usability!"?

--
Sergej Malinovski [http://dreamer.nitro.dk]

[prev in list] [next in list] [prev in thread] [next in thread]