Re: [kidsgames] Generic adventure game engine

[Manuel Gutierrez Algaba <irmina@ctv.es>]

Right now, I'm not very free for discussion/programming, moreover
I'm a bit upset with "Open Source movement", I'm the author
of many projects interesting from a technical point of view but
with a complete lack of support from anybody (see URL's below).
Anyway, I may boast that I have a long "career" in this world 
and I've learned a lot from continued "flops", bugs and failures. 
Anyway I've thought many times about these engines and I can 
say a couple of things.

> 
> Yes HTML seems to be the "best" way to do this... Plus it's quite 
> cross-platform

Ok, HTML is by far the most portable language. And it behaves 
wonderfully well, unless you need complex drawings/pictures with
animations. Generating a tar ball with 13000 pages of .html comprising
an adventure is very easy and portable. Each page can have some 
fork-points to derive and live the "story" . You can produce .jpg
so the story is "graphical". This method can be used to very
simple games like : guessing letters, guessing numbers.... but little
else. 

The advantage of this method is that you can program the engine
quite easily and you have freedom to use whatever .jpg you like. 

An engine of this kind could be made of :
- a list of specifications of relationships 
- a translator of those specifications into objects
- a HTML renderer with combinations , forks, and dumb stuff

For this python or tcl would  behave superbly. 

> Animated Gif are a bit limited, but DHTML is quite powerful to so simple 
> animations:

DHTML always sounds to me to Micro$oft (?) and modern browsers 
(not available for everybody).

> I like the fact that it's vector oriented and XML

XML is very fine, but who is going to mess to write it? We're talking
about non programmers?

> 
> I guess that's the good thing about "educationnal" games: the character is 
> important, in order to present you the educative content. It's like a 
> "standard interface" where you feel at home. Tux might be a good character 
> for that...
> 

Well, I wouldn't disneyize everything, we can spare having Tux
_everywhere_ .

> community. Like you said, it's text only (well the new engine allows images 
> and sound)
> Text has a limited "educationnal" range, (because it's NOT attractive at 
> first) but can be a way to have children read.
> Since text is still present on the internet (with ICQ, IRC and the like), 
> may be some interactive fiction could help to "discover" books: most of the 
> current IF deals with fantasy, aka Tolkien, but there are some Alice in 
> Wonderland stories... I'm sure it could be used to create Shakespeare worlds.
> The engine has been extended/recreated. There is a long spec here:
> http://www.gnelson.demon.co.uk/zspec/
> and the rest of the site is very informative on the Z-engine...
> http://www.gnelson.demon.co.uk/inform.html
> 

The question is what kind of games are we talking about:
- a large range of adventure of the "role-kind"
- simple games to learn simple things with little interaction
- a mix of arcade (shoot'em/choose'em all) arcade and "heavy-thinking"
decision-making games. 

> >Maybe you could write a specification?
> 

If you want to do something powerful and that covers many fields,
it may be complex to use and it may require many external things. 

If you want to focus in a limited problem you'll perform better,
but you'll get less support (I'm thinking about my Crossword stuff ,
for example). 


> I'm thinking about it.
>   Right now, I'm looking what and how the previous engines work. If you 
> have any "experience" playing with them, or with any "educationnal" games, 
> or have the need for an educational game, feel free to add your point of view
> 
> I think there is a need for a "click and write" game creator,

A Visual Free Game Designer... it sounds well, but it requires 
lots of programming, choose the right languages, and....SUPPORT. 
You may end with an interesting product but that nobody uses or
with little feedback. 
Thinking about it:
- I'd choose Tk for the interface.
- I'd design a couple of languages:
  * A "layout" language
  * An "action" language
  * An "actors" language
- Some renderers:
  * A Java one
  * A Html one 
- The internals of those renderers and interpreters of those language
should be done in python . Using Jpython the Java-rederer could
be very easy. 

> 
> >Is Python really that good (or really that easy)?
> 
> Good ? depends on which point of view !

Python is bad from the point of view of Perlers and from the point
of people who doesn't know it, yet!

> Well, it's quite easy to learn (but I'm not a good judge) and object oriented.
> Plus there are a lot of "educationnal" people around python (cf CP4E and 
> other python project here: http://www.python.org/sigs/edu-sig/ or 
> http://www.python.org/cp4e/)

> Well, this is the "gnome" way to do components... (KDE has another 
> solution, and Mozilla another one... And Microsoft another one too)
> intro on Bonobo: http://developer.gnome.org/doc/guides/corba/
> intro on Corba: http://www.javacommerce.com/CorbaNotes/index.htm

Corba sounds pretty well, how is going to use so much technology
for this issue (which is not hot, like Licq, gnome).

Educational soft is not very interesting for the average geek, or
hacker or whatever they name themselves. 

Regards/Saludos
Manolo
www.ctv.es/USERS/irmina    /TeEncontreX.html   /texpython.htm
/pyttex.htm /cruo/cruolinux.htm ICQ:77697936 (sirve el ICQ para algo?)

  QOTD: "I'll listen to reason when it comes out on CD."


-
kidsgames@smluc.org  -- To get off this list send "unsubscribe kidsgames"
in the body of a message to majordomo@smluc.org