GNOME.org
 

Some criticisms of GNOME


Hi all,
Here's an e-mail I got from the editor of tuxmag, which details some criticisms of the desktop. It raises some points that are interesting, and to which we should probably have an answer. 

Cheers,
Dave.

Tux Editor wrote:

David,

The numbers come from Evans Data Corp.  And it's a no-brainer to see
that the numbers reflect how real people think.  Without ANY help from
me, my 11 year old daughter jumped right into KDE and had no problems at
all using it and customizing it, right down to the way the panel looks
and works.  She hates GNOME with a passion, partly because it is a total
enigma to her, and partly because it's so darned ugly.  Okay, that's
just a matter of taste, but I happen to think it's ugly, too. 
Mango rips GNOME a new one in the next issue for many of the reasons my
daughter hates it.  I used to have fun with venom, but now that Mango
provides that style I let her do it and I can take a different approach
to my columns.  But even if I'm kinder and gentler these days
(sometimes, anyway), I must agree with Mango on some things.  For
example, I totally agree with her upcoming statement that the file
open/save dialog (I call it the file picker) is worse than bamboo shoots
under fingernails.  I can't imagine anything less intuitive than a
GNOME/GTK file open/save dialog.  That's one of the things that really
made my daughter hate GNOME and GTK-based apps that use that dialog. She gets totally lost when she has to deal with that dialog.
The GNOME interface is also inconsistent in the way it handles things. I won't go into detail now, but GNOME often makes things easy for a user 
at the cost of limiting what GNOME can do afterward.  Mango hints about
one of those cases, so read her column if you want an example.  Her
example also points out that KDE is too difficult in some ways, too, and
I agree with her 100%.  KDE is far from perfect.

I think the spacial Nautilus is nuts, and this is coming from a person
who used to love OS/2 -- and the OS/2 workplace shell worked almost the
same way as Nautilus works now.  I can deal with spacial Nautilus, but
IMO the problem isn't the concept.  The problem is that the people who
went with it jumped into it too quickly.  They didn't think it through
and provide options for those who wouldn't like it the way THEY liked
it.  For example, what about those users who don't want to keep opening
new windows on the desktop?   Yes, I know you can FINALLY use a GUI way
to change this behavior NOW.  But when spacial Nautilus was introduced,
the only way to change the default behavior was to change a registry
setting. Now THAT is a total lack of foresight. 
But wait, there's more.  Right-click to use the browser mode?  Totally
unintuitive. Windows makes that mistake, too. 
But wait, there's even more.  Shift-double-click to close the previous
window?  How intuitive is that?!?!?  Why not simply provide an OBVIOUS
global option (a checkbox in an obvious place) that tells Nautilus to
close the previous window when you navigate to another folder?  If I
recall correctly, even the OS/2 designers provided that option.  Problem
solved -- all it took was a little forethought, which is something the
GNOME developers totally lack. 
Speaking of which, I don't know if you still can't use the shift key to
close the previous window when you set GNOME to open things with a
single-click, but that's yet another example of GNOME developers lacking
forethought.  Some people like to single-click things to activate them,
and GNOME lets you switch to single-click.  Yet you couldn't
shift-single-click a folder to open a new one and close the previous
one. Didn't anyone consider that people work differently than they do? That's just really bad QA. 

But what probably irks me the most about GNOME is that it forces you to
choose between what OTHERS have decided your desktop should look like. This is the same as "Didn't anyone consider that people work differently 
than they do?"

You have a tiny bit of tweaking room (you can mix and match pre-defined
icons with pre-defined window styles and pre-defined widget styles), but
you can't do something as simple as pick the color of window title bar. I've heard GNOMEies and GTKies say that this is deliberate design 
decision.  It keeps people from doing something stupid like making the
window title bar white and the text white (and therefore make the window
title text unreadable).   I'll believe that excuse when I believe in the
easter bunny.  Face it.  GTK simply wasn't built to let normal humans
customize things like the color of the window title bar, and it would be
a bear to go back and re-write GTK to work that way.  I suspect nobody
wants to do that, so it stays the way it is, and people keep relying on
the excuses for the dumb behavior.  I'd be much more kind to the
GNOME/GTK authors if they'd just be honest and admit they screwed up and
had no real foresight when they built GTK.  The exuse is lame.  These
are computers, after all.  Computers are perfectly capable of comparing
color values.  If the programmers REALLY were concerned that someone
might make the window title text unreadable, the RIGHT solution would be
to put safety checks in the color customization program.  It's not that
hard to write a program to check the color values the user is picking (I
don't mean the numbers, but the values - an artist would understand) and
warn the user that they are about to make the text difficult or
impossible to read.  Then provide an "undo" button that isn't affected
by the color changes.  They see their mistake, hit "undo".  Problem
solved without forcing users to do things YOUR way.

You said, "I fear that there are some fundamental differences between
the way the GNOME community would like our desktop the be, and the way
you would like the desktop to be, so I am not sure that your opinion of
us will improve over the next year or two."

I think this gets right to the heart of the problem.  I don't want to be
a part of a GNOME community.  I don't want to be a part of a KDE
community.  Personally, I don't think there should be a GNOME community
or a KDE community.  That kind of thinking is self-centered, not
customer/user-centered.  Both GNOME and KDE developers should find out
what users want -- from the newbie users to the power users, and build a
desktop to please them all.  Provide an easy way to get to the power for
those who can use it, and hide that power for those that do not want to
be confused by it.  Linspire is a decent (not perfect but decent)
example.  It simplifies KDE enough so that people aren't confused by the
gazillion customization options. But it doesnt' DISABLE those options. A power user can use Linspire and tap into everything KDE has to offer 
if he/she wants to.

And who is the GNOME community, anyway?  The community of GNOME users or
the community of GNOME developers? I suspect most of what you call the
GNOME community are GNOME developers or contributors (or just people who
give advice on the GNOME lists).   If the GNOME community is building a
desktop to please themselves, then fine -- but the consequence of that
is that GNOME will always be a desktop for GNOME lovers, not for the
average desktop user.  If you don't mind having that kind of a limited
audience, then fine.

Okay, I'll get off my soapbox now, but I really think you GNOME guys
should re-think the way you're approaching the desktop and make some
MAJOR changes.  The first step I recommend is that you ditch that
completely absurd file picker and replace it with one that is BETTER
than the KDE file picker.  The KDE file picker is incredibly attractive
and powerful, but it has weaknesses -- you could kick it's butt if you
just put some thought into it. I have ideas if you're interested. 
Feel free to post this on any list you want.  I think GNOME developers
need a major kick in the keister if that's what it takes to learn that
people don't necessarily want what THEY want.

-Nick


--
Dave Neary
bolsh gimp org
Lyon, France
[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]