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

List:       kde-usability
Subject:    Re: System Settings as a Design Lesson
From:       Celeste Lyn Paul <celeste () kde ! org>
Date:       2008-12-02 17:27:56
Message-ID: 200812021227.56573.celeste () kde ! org
[Download RAW message or body]

On Tuesday 02 December 2008 11:45:02 Rafael Fernández López wrote:
> I am the culprit of this change.

oops :)

> About what Celeste pointed out, I could agree a bit more (even in the
> disagreeing rank) if she had painted each "category" alternating
> backgrounds (as Mac OS X does).

It doesn't matter if the border is changed to the "outside of the box" or the 
background color alternates.  It it has the same effect and I'm OK with either 
one.  

I have a problem with the current implementation because it is less efficient 
than the old one. In order to offset the border-under-header effect, you must 
add more whitespace to create an object area. 

 In Dolphin, there is extra white space between each category. In System 
Settings, it is barely perceptible. If the space were added to correct the 
design, then you would be wasting a lot of space in a UI that is already at 
risk for scrolling.

> Not even talking about Dolphin. For me the explanation followed has no
> application at all in Dolphin.
>
> Note that I am not a usability expert. I am talking from my personal
> opinion as a developer and as a user. Note that I am open mind, I read, I
> understand and I can change my opinion. Note that I can implement something
> I am personally against [better said: that I don't support] if it benefits
> more people that it "damages".

It doesn't damage anything except for people who absolutely cannot stand 
change. If anything, people were more affected by the fact that the screenshot 
was in qt3 and not pretty, or that the line in the kde4 version had a nice 
fade effect. 

People have this misinterpretation of usability == preference, which it is 
not.  A small part of usability considers satisfaction because given two 
similar designs, users will perform better with one they they like and find 
more aesthetically pleasing.

For your reference, usability is:
* Learnability
* Efficiency
* Memorability
* Error prevention
* Satisfaction

Very much functional design and none of this marketeering hogwash for 
preference.

> What I wonder now. So, from your point of view, Dolphin should also use
> this layout ? (the one you propose)

If one uses it, they all should so the environment is consistent.  It would 
also make it visually consistent with how Plasma is promoting the use of 
section headers instead of group boxes. This change would conceptually alter 
the line from an embellishment to a border.

> Do you think that this will work that well when I set a bold font as my
> default font ? It could seem so much a "whole block" that it could confuse,
> IMHO.

This would be a problem besides this view.  Plasma is currently using and 
promoting the use of bold fonts to replace group boxes. We could consider 
making all headers larger than the default font size so it is double coded 
(size + weight) for normal default layouts and still single coded (size) when 
you make all of your text bold.

Also, the border will just be on the top, not the bottom, so the header is 
still anchored to the border which will make it stand out.  The only difference 
is that the header will look like it belongs to the contents more than it does 
with the line separating it.

Can you make the change and send a screenshot of Dolphin?  It might be easier 
to talk about if there is a picture to look at.

> Is strange to me too because I can see everywhere documents where the title
> that "sums up" or "gives you an idea" of what contents follow usually are
> underlined, or even underlined till the end of the paper (as done here, but
> on a screen). Does this still apply on those cases ?

Do you have some examples outside of print?

> Please, please, please... don't take this message as "free" criticism. As I
> said I am open to all ideas [I am even rewriting KCategorizedView because
> of performance issues and I could take this in count, or for systemsettings
> create a special view], and I am not wanting to do any kind of flamewar
> here. I just want to learn.

This is really a minor MINOR issue compared to all the other System Settings 
problems, such as a bad information architecture (what I plan on talking about 
next). It is discouraging that a simple discussion meant to be educational on 
a basic design principle gets everyone so upset. At the same time, it is also 
a very basic error that I would not hesitate to point out if it were a client 
application and would expect them to fix it.

-- 
Celeste Lyn Paul
KDE Usability Project
usability.kde.org
_______________________________________________
kde-usability mailing list
kde-usability@kde.org
https://mail.kde.org/mailman/listinfo/kde-usability

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