From: "Hal E. Fulton" <hal9000@...>
Date: 2001-08-16T05:34:07+09:00
Subject: [ruby-talk:19842] Re: Ruby GUI

----- Original Message -----
From: <avdi@avdi.org>
To: ruby-talk ML <ruby-talk@ruby-lang.org>
Sent: Wednesday, August 15, 2001 2:35 PM
Subject: [ruby-talk:19829] Re: Ruby GUI


> Quoting "Hal E. Fulton" <hal9000@hypermetrics.com>:
>
> > The concept of a new GUI is somewhat appealing,
> > though I'm not full of ideas for it.
>
> I have a few, but I'm just a 21-year old with delusions of grandeur... ;-)

Ha... don't worry, when you're twice as old, you'll have
twice as many delusions.... ;)

[snip]

> The people at the Berlin project seem to have a lot of neat ideas:
> http://www.berlin-consortium.org/. Also, whoever perpetrated the Morphic
> framework on Squeak probably has a lot of worthwhile things to say on user
> interface design.

I'll check that out...

[snip]

> I think what might be necessary for a successful toolkit is to have a
number of
> layers, where each layer is a different tradeoff between control and
> abtraction. At one end of the continuum, you could simply express your
intent
> in an XML interface file - i.e., "The user needs to be able to choose
between
> values x, y, and z for the user option 'foo'" - and the UI system would
> generate a command line option "--foo=", a GUI menu option "Options->Foo"
and
> an interactive command-line prompt: "Whould you like 'foo' to be x, y, or
z?
> (x)" as appropriate. At the other end of the continuum, the interface
> definition would specify the precise x/y coords and size of every widget,
and
> would either explicitly or implicitly declare the UI capabilities that
/must/
> be available on the system for the application to be run.

That's roughly what I'm thinking.

> > I also have some widget ideas that I think
> > are pretty good but are untested on humans. :)
>
> Are you saying you practice animal widget testing? You barbarian! ;-)

My cat swears she doesn't mind.  :)

Hal