[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Desktop normalization

I was deep in meditation when Marcin Krol awoke me by saying:
> For heaven's sakes, no! Even bad standard is better than lack of standard.

I don't think so. With a lack of a standard, one has the option of weeding
through available products/implementations for a "best-of-breed" product.
Usually, a de facto standard will emerge this way. With a bad standard, one
has only the option of using a bad product. This is exactly what has turned
many away from Microsoft and Microsoftesque (new word ;-) products, that
there is only one product to turn to and the product is usually based upon
a faulty core.

> >  I think the whole KDE vs Gnome thing is total bullsh__.
> It is, but not for reasons you state below. Just why should someone
> split because of some fifth-grade importance thing in license is 
> beyond me. OSS is no better at avoiding unnecessary redundant
> effort than commercial software.

I don't use KDE because I like fvwm2 (and I've only begun to tinker with
Enlightenment). At school, mwm is the default (soon to be mwm w/CDE). I
had to change my wm there to fvwm2. I think this is what gives X its
greatest strength, that I can choose a completely different look and feel
by simply plugging in a new window manager. The idea that people should be
able to customize more than just their screensaver and desktop colors is
one not to take lightly. Unix uses are used to customizing everything,
from their emacs modules to their shell aliases/functions. I don't 
personally have anything against KDE or Qt (which is now "open", btw),
but I like an environment which gives me the control of choice. So, I've
chosen GNOME, because it is window-manager independent. It's still in
development, but then again, so is the kernel I'm using ;-)

> >  As I
> > see it, Enlightenment and GNUStep (Windowmaker/Afterstep) are at least
> > _AS_ popular (me personally perfering a total "Stepified" desktop,
> > running Windowmaker, patching GTKstep into GTK, etc...) and believeing X
> > has to fix itself (somewhat in agreement with Jim Gettys (
> > http://editorials.freshmeat.net/jim981031/ ).
> Probably E and GNUStep are. However, it is not important which desktop is
> most popular *right now*.  KDE would be very important even if it were not
> very ergonomic or aesthetic. It is important because it is first attempt
> to create extensive, complete desktop for Unix. Not only look and feel,
> but also basic desktop, everyday apps. Standard, available everywhere.
> Don't want them - install something better. But *something standardized*
> has to be commodity. 

Standards should never attempt to define themselves. Standards should be
codifications of de-facto standards, much like law is often the codification of
commonlaw. I don't think the purpose of LSB is to steer the course of Linux
or Linux environments. I think it is to codify and unite the many similar but
yet unique aspects of Linux. Where two disparate technologies clash but need
to be standardized, then a means of cooperation has to be developed. I hesitate
to say which environment or window manager is in greater usage, and to base
any standard upon one which is not (or just barely is) is to upheave the
entire process and at the same time lose credibility where it counts:
the users. What does a standard based, for instance, on KDE mean if only
30% of users use it? A company which expects KDE will be encouraged to
write KDE-only features into their product, because they will falsely
believe that everyone can handle that functionality. In fact, they will
not do very well in the market and will gain a distaste for the platform
in general. 

Clearly there needs to be a standard for GUIs. As Alan Cox wrote, that
standard is X11 (for now). If a developer writes for X11, they will be
sure to have a program compatible with almost all Linux users (who use a
GUI). I will go a little further to say that GTK should also be included
as a standard toolkit, only because the de facto standard, Motif, is
commercial-ware and wouldn't fit in the standard at all. That's not to say
that a developer couldn't develop for Motif, but they will be aware that
GTK is gauranteed to be available, while Motif is not. I don't think the
end result is that much different (except that GTK is [supposedly] themed).

> > Of course this is all just my opinion, and not a very important one
> > really.
> Of course you are entitled to your opinion. Everyone is. But also there is
> this problem of standardization. Having variety of window managers on X is
> like burning the house to roast a chicken: you gain something but loose *a
> lot*. 

No, it's like being able to choose what color your house is, the furniture
in it, where the furniture goes and what you watch on the TV of your choice.
These are good things, IMHO.

> > I think the LSB should be conserned underlying structure (file
> > structure, sysV vs bsd, basic libraries maybe, default applications like
> > ls, cd, rm, mv, cat, etc...), on which larger software componants (like
> > a windowmanager) can work out thier compatibility problems themselfs.
> But they *won't* work out their compatibility problems - they did not do
> it so far, what makes you think that teams will magically start weeding
> out incompatibilities overnight? You seem to hope that *finally* these
> cats will herd themselves some time. Looking at history, it did not
> happen, and those who do not know history... 

I don't think the LSB should standardize something simply because no one
will do it. These groups haven't worked out their differences because they
are preliminary and experimental. Eventually, one will become dominant and
the other will be made compatible or will become obsolete. This explains why
most window managers offer some means of understanding Motif "hints". Motif
is dominant, other products must accomodate. That's the way the free market
works and I would hesitate to attempt to suppress this tendancy. Again,
I think LSB will be much more successful if it sticks to a role of
codification primarily and specification only when absolutely necessary.

> > An API to access either Gnome or KDE desktop strikes me as beyond the
> > scope of the LSB.
> 1. How about *generic desktop API*. 2. If there is any reason for LSB, it
> is desktop. 
> Desktop is everything.  Who controls desktop, controls computing. Has
> success of MS taught unixers nothing? "Alice in Unixland" should be
> required reading on unix courses.

I don't know about this. When someone with Slackware tries to download
a glibc Netscape, what their titlebar looks like isn't going to be the
first of their worries. Binary compatibility is the first and foremost
issue. Desktop is an issue, but you know, any good desktop will be 
reasonably easy to understand. Bad desktops won't be used and so won't
matter. I think if we leave it to the free market, in the long run we
will achieve a better product.


| Aaron Gaudio                   mailto:icy_manipulator@mindless.com |
|                    http://www.rit.edu/~adg1653/                    |
|      "The fool finds ignorance all around him.                     |
|          The wise man finds ignorance within."                     |

Use of any of my email addresses is subject to the terms found at
http://www.rit.edu/~adg1653/email.shtml. By using any of my addresses, you
agree to be bound by the terms therein.

Reply to: