Re: A Violent Realisation [Was: Preferences]

[Miguel de Icaza <miguel ximian com>]

> Miguel de Icaza <miguel ximian com> writes: 
> > I want to get work done just as much as you do, but reading this thread
> > is scaring me.  It scares me because I am thinking `I am being
> > marginalized, and my feature I depend on will not be there, and I will
> > end up running some other desktop that does do what I need'.
> 
> The goal of the project should be to appeal to users who did not like
> and could not use the traditional desktop.

Well, that is one goal.  Different people contributing to Gnome have
different incentives to do so.  One very valid argument is `MacOS X and
XP, KDE do not cut it for me'.  But some others are might be thinking
`OS X and XP cut it, but they are not free'.

So the goals of the project are different depending who you ask.  You
have stated your vision.  For example, I think that enabling the open
source desktop on the enterprise is a good strategy for open source, and
hence, I am more interested in that scenario, and not too interested in
providing a desktop for people who `did not like traditional desktop'. 
Maybe they like it, I see no reason why we have to deviate from that. 

> I think we can do that without losing traditional free software
> users. For example, KDE has many such users and doesn't have viewports
> and never has. ;-) But yes, the traditional free software users are
> not and shouldn't be our _primary_ focus.

You are putting too many people in the same bin.  Who is your
traditional free software user?  

People running production servers range from the enthusiast, to people
who are just running Linux servers as a job.

> But Miguel, you _also_ haven't answered the questions I've posed:
> 
>  - what is the guideline on which prefs to include?
> 
>  - how would you refute or address the argument that each preference
>    has a cost? If you want to make a convincing argument, look at the
>    specific list of costs in my little essay, and address each one.

I joined this thread late, with Jeff's last quote.  I will go back and
read your essay and address some of those issues.

That being said: I did like your essay when it came out, and I linked to
it from my blog as a `really nice article'.  

We have too many options and that leads to bugs, usability problems and
related problems.  I agree that too many options are wrong, and I agree
with Joel that many times the option is there to avoid having to make a
decision in behalf of the users. 

But program options have a genuine place in the world.  Please follow me
along this example.  This is just an hypothetical example to illustrate
my point, I am not advocating anything like this:

    It could be argued that we should make a decision to have the panel
    only live in the bottom of the screen, because that will avoid
    confusing users, or will remove some option from a menu, will kill
    some extra code in to be maintained.  You name it.  A decision is
    made to only support a panel in the bottom of the screen.
    
    The problem with this scenario is that this is assuming that
    everyone works the same way or deals with the same kind of software
    applications that lend themselves to be used in that way.  Some
    people might need more vertical space, some people might need more
    horizontal space to run their applications (ie, not browsers or
    xterms).
    
    Now, even apple supports moving their panel, indeed it is one of the
    much trumpeted new features of their OS upgrade.  And the reason is
    simple: various people needed different setups.  They also support
    auto-hiding their panel.
    
    Sometimes it could be a matter of need (the application demands
    require it) or personal choice.  Just like you choose the color of
    your car or the clothing you wear every day.
    
Anyways, that was just a sample.  What I am afraid of is that this will
lead to a quest to remove every option without having everything in the
balance.  Yes, some apps have too many options (xchat) and making those
better is a worthy goal, but this should not become a crusade to remove
options. 

Lets examine another case: palette configuration for a terminal
application.  Although you could allow the user to select every one of
the 16 colors the application supports individually that would be a poor
UI choice.  I like what we did (I dont know where the idea was copied
from originally) which was to have different "terminal palettes": the
linux console profile, the xterm profile, the rxvt profile.

Miguel.



> This thread is just a lot of pointless emotional speculation based on
> unfounded fears, unless those questions are addressed.
> 
> Havoc
>