Re: Owen Taylor's paper

On Tue, 17 Sep 2002 00:11:48 -0400 (EDT), you wrote:

>On Tue, 17 Sep 2002, Gerald Henriksen wrote:
>
>>However what Red Hat is attempting is doomed to failure as long
>>as Gnome and KDE insist on increasing the differences between
>>the 2 projects (and hence the applications written for either of
>>them).
>
>If that holds true, then Linux itself as a viable desktop 
>platform is also doomed to failure.

Well I don't think so, and I certaily wouldn't be starting to write a
desktop app with my free time if I thought nobody would be using it.

However unless both the Gnome and KDE projects suddenly make up and
become friends the reality of the Linux desktop is that it will
actually be 2 products, Linux/KDE and Linux/Gnome, because their UI
design is almost exactly opposite.

Consistency is very important to large companies.  Prior to becoming
unemployed I had to commute to work on the train and a benefit of that
was talking to people who developed software for large companies and
provided support.  Consistency, either in hardware used company wide
or in the interface design for the software was extremely important.
You screw up the consistency and you see your support costs go up as
more users call in confused.

Presenting a KDE desktop to a prospective customer and saying that
Evolution is your chosen email client and mention that by the way in 6
months or so when Evolution gets upgraded to Gnome2 standards the UI
design of that app will be the exact opposite of what it currently is,
and the exact opposite of everything else in KDE, and you will be
laughed out of the room.  Few if any companies will want that support
nightmare.

Presenting a Gnome desktop to a prospective customer and show that
while Gnome uses one UI, Mozilla and Acrobat Reader use another, and
the clients in house Java apps are consistent with Mozilla and Acrobat
but opposite of Gnome, and again you will likely be laughed out of the
room.

Any Linux desktop that will be acceptable in terms of minimal
retraining costs and minimal long term support costs (including less
stress to the workers using it who after all are only using because it
is part of their job and not because they like computers) will need to
maintain a consistent UI across both the desktop and the applications
in use.

While I certainly appreciate the time, effort, and money that has been
put into the new icons, menus and themes by Red Hat and the people who
actually have done it I can't help but think it is like putting a new
coat of paint on a sinking boat.  Yes the boat will look really nice
but it is still sinking to the bottom.  All Red Hat is done is put a
new coat of paint on while ignoring the fundamental UI problem
underneath.

Of course the ultimate irony is that it is the Red Hat backed Gnome
project that has caused this UI problem for Red Hat by choosing a UI
standard that is opposite of everything else in use on Red Hat Linux
(Motif, KDE, Java, and Windows apps under Wine).