>Alain: What a really hot idea that could be fruitfully
>expanded upon. An engin that employs heuristics in the
>form of editable rules, to detect virus-like activity
>for example. The same inference engin could also be
>used though for other interesting meta-level features.
>My favorite plan in this regard is to create an
>agent-based system that models its own behaviour ->
>A-life.
Alain,
could you tell me what A-Life is in a few words?
>Alain: What are we doing, conceptually or practically,
>to make FreeCard innovative? Are we envisioning a mere
>clone of circa 1987 software, or (ultimately)
>something that will have an edge that xCards haven't
>touched yet?
1.0 will be just HyperCard 2.4.1 with the editing environment being
implemented as stacks (no use in adding a built-in editor, stacks will be
much easier to maintain and won't require any porting between platforms).
It will also internally remove those darned limits and thus be able to
handle color graphics, though we might not yet add commands for handling
color yet (the edit menu commands will very likely already work for color);
and it will use system-native look-and-feel.
At least that's what I think is what we agreed upon and what is
technically necessary. I think it wouldn't be too smart to add actual
features for 1.0, as this could lead to feature-crazedness at the loss of
the original HC feature set. By strictly sticking to HC's feature set for
1.0 we will make sure everybody gets at least HyperCard.
For subsequent releases my major thought would be to do the same Apple's
developers are doing with Carbon: Listen and implement. I.e. hear what
features scripters request the most and add those. I guess this will be
when we add user properties (if we don't need them already for the editing
environment) and start peeking at other xTalks to adopt ways to satisfy
scriptsers' requests.
In the very long run (2.0, 3.0 ...) I'd love to see web deployment both as
a server-side or client-side file (i.e. FreeCard stacks as web sites and
FreeCard stacks as FTP programs, to use *greatly* simplified examples) and
addition of more OOP-like features (a concept similar to classes would be
cool).
Anybody have a different vision or would like to add something?
Cheers,
-- M. Uli Kusterer
------------------------------------------------------------
http://www.weblayout.com/witness
'The Witnesses of TeachText are everywhere...'
