[Kde-games-devel] compiling kblocks

[Mauricio Piacentini]

(same here, forgot to reply all to the list)

Dapeng Zhang wrote:
 > The current status of the program is fine for me. I would create a 
"demo" entry and hook to the game as soon as possible. Is it possible to 
make a neat interface with only virtual functions for
 >
 >      controling the block and accessing the status of the game
 >
 > I will try to define this interface in next a few days.
 >
 > What do you think?
 > We can see in two months whether or not it is possible to make a 
human v.s. AI competition.

To be honest, the AI or human vs computer functionality was not 
something I was planning for KBlocks. My goal was to produce a leaner 
game, with only single player functionality, and highly configurable 
(visually) via themes. And also use it as a testing ground for 
technologies like KNewStuff, Phonon and SVG-animation integrated with 
QGraphicsView (see the concept of BlockAnimator objects, for example.)

You can see I was more concerned with what could not be done (easily) by 
extending ksirtet, so basically the development of KBlocks was justified 
by having a new design that is centered on eye-candy, basically. :)

As I said, I can see how the AI can be useful for the demo (for 
example), and how better organization of the data structures internally 
can help. I just do not want to make it more complex than necessary, as 
this has costs in the future for maintenance and upgrades. And I do not 
want to do this work right now, until I complete my initial design, 
which was centered around themes and how the game looks and plays for a 
single player experience.

On the other hand, ksirtet is already mature in terms of data 
structures, and already has an AI and human vs computer modes. So for 
research, it might be better to check it out from SVN (see 
unmaintained/4), ressurect it to playground (it builds with 4.0 afaik), 
and experiment with the replacement of the existing AI, maybe. You would 
have all the hooks and functions for AI already in, and even an AI 
implementation that you could use to compare against yours. And maybe 
pitch them against each other, which could be cool to see!

So two courses of action: you can work on extending the unfinished 
KBlocks and we will see how it goes (at least for the demo, this could 
be wonderful.) Or you could checkout the last version of ksirtet from 
playground (this is revision 688508), and see if the AI interface 
defined works, and maybe even use it as a starting point. The only 
reason ksirtet is not in playground is because no one was working on it 
actively, it might come back if you (or someone else) volunteer for this.

Regards,
Mauricio Piacentini