Re: the same page

[Dick Porter <dick ximian com>]

On Thu, Dec 20, 2001 at 12:41:43PM -0800, Seth Nickell wrote:
> > Frankly, the market of ex-windows users (or 3d graphics and CAD users) is
> > totally insignificant compared to the market of people that have never used
> > a computer before.  Bearing this in mind, and the fact that the evidence
> > seems to be that once someone has chosen a particular system they are
> > very unlikely to change to a different one, GNOME as a project should aim
> > to be that first choice.  With the reality that the wealthy parts of the
> > world have made their choice, and the poor parts havent, then it seems
> > obvious (to me at least :) ) where we should be going.
> 
> Which is why I make the point that Windows is not the prohibitively
> expensive part of the computer. If Windows cost three times what the
> hardware costs, maybe we would have a huge oppurtunity here. But unless
> GNOME is a *LOT* better than it is now, we're toast, because Windows95
> is a more usable system than GNOME right now. We shouldn't slow down our
> growth in features to match slower markets yet because we simply aren't
> to a point where people can use GNOME. Esp. people in areas that haven't
> used computers.

I think we've a fundamental disagreement over the word "better".  You equate
it to "more features", I want "more efficient use of what we have first".

And I don't just mean "more efficient" to be "use less disk and RAM" - I
also mean more efficient to use.  For example, my suggestions a long time
ago that the control centre should have a set of task-based setup paths for
some common items (I think the example I gave was setting up a palmpilot
to sync with whatever.)

> > 
> > And note than I am emphatically _not_ saying that we shouldnt exploit new
> > hardware power.  What I _am_ saying is that we shouldn't insist on it, but
> > that we should just use it if available.
> 
> That's really really hard. If you buy that its the layers of abstraction
> that are causing a lot of the problem, it doesn't matter too much if you
> are using a non-roman alphabet or not, you're still going to be paying
> (at least some) for pango.

That depends on the implementation.  Is it possible to build Gtk 2 without
pango support at all?  If so, shipping a set of GNOME2/lightweight packages
with default configs set to "minimalist" would be one possible answer.

> 
> -Seth

- Dick