Desktop normalization

Marcin Krol lug at
Tue Nov 24 02:24:11 PST 1998

On Mon, 23 Nov 1998, Rob Current wrote:

> It's NOT. KDE is ugly as sin.

A matter of taste in 100%. 

> placed which included such things as:
> NAME      Large Icon Small Icon         Command Line
> MAIN--
> Netscape netscape.xpm netscape-mini.xpm /usr/local/netscape/netscape
> EMACS    emacs.xpm    emacs-mini.xpm    /usr/bin/emacs

Terrific! That's all I ask for. Somebody in response to Jim's editorial in
freshmeat has written that there's a need for something like User
Interface Markup Language, a thing that in my understanding would relate
to desktop like HTML relates to Web. Even though there are netscapisms and
explorerisms in existing Web pages, they still are mostly usable without
having to tweak for 20 hours in the browser.  That's the goal. 

> etc.. Now, little shit like that would be a rocking idea if standardized,
> and slowly all existing windowmanagers could evolve to read the standard
> menu structure that was simple, easy to edit, and be universally read in
> the end.

> But, beyond that, the Window Manager "race" is something that is a
> strength of GNU software in general, and to pick one before any of them
> are mature (with the exception of fvwm and maybe *step), would be a big
> misake.
> X itself has a ways to go, like it or not, user interface is a very
> personal thing, a rapidly developing thing, and something that "isn't
> there yet."

> If one UI was for everyone, people would have never bought Macs,

I thought that *one* UI is the reason why people bought Macs.

> If UI
> mattered to evryone, no one would have ran UNIX over the last 20 years.

Yes and no. Take note of several things:  
- availability and price/performance of GUI terminals in the time of the
most explosive UNIX growth
- characteristics of typical user in various environments (niche,
education, skills, etc.) 

> Starting to lay out some X standards to which all WindowManagers could
> start to conform is a nobal idea.  But, just how much do you want to have
> a LSB?  I mean, it's going to hard enough to get the base down, and keep
> it maintained (as libs versions get newer and basic applications get
> better).
> CDE as the standard, and how that all came about could have killed all of
> UNIX, let's not do that... 

> Maybe if you feel strongly about it, work
> on/with something to do with a "XSB" where X is the focus, not Linux.

Maybe you are right. I'll think about it. I do not really care (and
I think users do not really care) if the goal is achieved
within this or that initiative, as long as it is achieved at all.  

> Believe it or not, Linux has a large user base who run "headless" boxes,
> and just don't care about X at all.

In that case standard for X environment does not bother them the
least bit, does it?

> I dunno, I am going to shut up now.
> Ok, done now. (^^But think, that is what emotions your going to envoke in
> millions of people if you choose to pick and push some standard X).

Pushing some standard by excluding others would be stupid thing
definitely. However, it does not have to be that way. All I ask for is
rather some default that works out of box, without inherent *need* for any
customization, but with *capability* for customization, so various
developers would be pressed to make software that works out of box, but
does not force user to make it only one way (MS approach), because
that way developed product would risk loosing too many users. IOW,
variety would keep developers on their toes to provide universal
product, but standard would also press them to make instantly usable

 Marcin Krol

Hiroshima 45                   Tschernobyl 86                      Windows 95

More information about the lsb-discuss mailing list