Re: swallowed apps

[Havoc Pennington <hp redhat com>]

Robert Hardy <rhardy webcon net> writes: 
> I must admit I'm extremely disappointed at the removal of swallowed apps in
> Gnome 2's panels. You seem to have axed a very useful feature (granted one
> which badly needed repair) for the heck of it.

It wasn't for the heck of it, it's because it didn't work (and could
not reliably be made to work). There's a protocol (XEMBED) to make
swallowing reliable with both swallower and swallowee cooperating, but
even that is not 100% reliable without an X server extension.

So there would be a guaranteed, impossible to ever fix, stream of bug
reports on the feature. Yay.

Plus, when asked, very few people were using it.

That said, anyone could write an "app swallowing applet" that would 
implement things the old way. It's most likely not a very challenging
programming project even.

> I'm sure it wasn't intended
> but this reeks of the "We know what the users really want" attitude of
> certain unnamed monopolies. I don't know about you but choice was one of the
> things that first attracted me to Linux, X, Gnome etc.

It isn't choice if your choices are:

 - bloated hard-to-use environment that has all features from
   every OS ever #1
 - bloated hard-to-use environment that has all features from
   every OS ever #2
 - bloated hard-to-use environment that has all features from
   every OS ever #3
 - bloated hard-to-use environment that has all features from
   every OS ever #4
 - bloated hard-to-use environment that has all features from
   every OS ever #5
 - bloated hard-to-use environment that has all features from
   every OS ever #6

It is real choice if your choices are:

 - several bloated all-features-ever environments as above
 - various simple environments that take different approaches

> If you insist on simplifying things to the point where they are
> dysfunctional for intelligent users in order to keep things "simple" feel
> free to do so BUT only do so in a novice/beginners mode. Don't totally
> remove the functionality and make things useless for those of us who are not
> challenged.

Talking about intelligence is wrong.  The UNIX/Linux desktop as it
currently stands is *still* too complicated for, say, many grad
students I've known. Are they stupid? no, they can tell you a lot
about astrophysics or whatever. Do they care about the computer as
more than a tool? no.

They just aren't interested in micromanaging the GUI, or having the
GUI be a construction kit for their own UI experiments.
 
> Zillions of options are only bad: 
> 1. if they are poorly presented
> 2. if they significantly affect performance to the point of unusability
> 3. if they confuse the user

As someone who's written many tens of thousands of lines of UI code
and watched the evolution/discussion of many more lines, I just don't
agree. My experience says otherwise. I've seen numerous prefs turn
into time sinks or UI disasters, and seen numerous prefs keep people
from fixing things to just work or keep them from fixing the defaults.
I'm not pulling this out of my ass, I've been hacking open source UI
code for years and I've seen the same patterns over and over.

Re: applets, they were not generally removed to reduce complexity, the
missing ones mostly vanished due to lack of a maintainer, and would
come back if someone stepped up to babysit them. But as long as no one
is going to do the work, the applets won't be there.

GNOME is very extensible, anyone can write an applet; you can
remote-control the window manager using libwnck (and quickly write
your own pager, window-matching type functionality, whatever you
like); you can load GTK modules that do various things.  It's all
about having volunteers.

If you can't code, helping out with bugzilla is a good way to free up
time for coders.

Havoc