Re: Writing on the wall

[Stephen White <spwhite chariot net au>]

On Mon, 31 Jul 2000, Mark Leisher wrote:

> So my last bit of advice is to keep your eye on projects that adhere to
> the philosophy of small and simple, such as FLTK.  They may seem limited
> compared to your current toolbox, but the new generation of programmers
> will dump Swing, Xt, Motif, Gtk/Gnome in a heartbeat if anything simpler
> approximates what they want.

That was an interesting point. In addenum, I'd like to point out that
there's no guessing which set of features is going to be required in the
future.

Projects get bulkier as they add enough of the most commonly requested
features to satisfy a greater majority of people. Up to a point, that's
good. It's nice to have everything in one place.

The best projects have the minimum number of features to attract the most
number of people. In other words, having all the features that most people
want, but no more. A project becomes bulky when it tries to have too many
features to make too many users happy. 30% effort for 70% of the results is
much better than 90% effort for 90% of the results.

The rot sets in when the existing features become old features. Maybe it was
once popular or it was a protocol that once looked hot but never caught on.
But there are existing users relying on that feature. Removing the feature
would piss off people far more than never offering it in the first place.

The only way FLTK can remain small is never to add new features, or they
lose users through API churning. That's why new products tend to catch on
faster than old projects that have been around forever. They offer the
right set of features within an up to date API that hasn't been twisted
around to support old features.

--
  spwhite@chariot.net.au