My entry to 3rd PGD challenge - no name yet

[pstudio]

Sounds like an interesting concept. I'm looking forward to see what you make.
You mention that you would like to create the map programticcaly. Does this mean that the map will be different for each playthrough or will it always be the same map layout?

[SilverWarior]

My intention was for having uniqe map every time.
But due to time constraints I probbably won't implement this into my entry. I just remembered how much time I spent working on this on my entry for previous challenge. And becouse it still isn't finished and would need athleast partial rewrite due to change of how I intent to manage map this time I think it is better for me to leave this out or I will run out of time again.

[WILL]

Procedurally generated maps would allow for nearly unlimited map layouts causing your game to have a ton of variety, however it would weigh in on your time to program as much. An alternate option is to create a set of pre-designed maps and randomly select one of those at game start to keep the game from playing the same almost every time.

I think it would be a good idea to have at least more than 1 predefined map layout in the game though. It would prevent the game from falling into the trap of having a "solve the puzzle" type of game play. It also would mix it up so your players aren't always trying the same stratgey so it has more replay value.

Predefined maps by nature do sort of fix your variables for balancing and benefits in strategy games generally, but with even a little bit of variety this can at least be diluted a bit.

[SilverWarior]

Procedurally generated maps would allow for nearly unlimited map layouts causing your game to have a ton of variety

Yes I know.

An alternate option is to create a set of pre-designed maps and randomly select one of those at game start to keep the game from playing the same almost every time.

I'll probbably go for this option.

I think it would be a good idea to have at least more than 1 predefined map layout in the game though. It would prevent the game from falling into the trap of having a "solve the puzzle" type of game play. It also would mix it up so your players aren't always trying the same stratgey so it has more replay value.

I'm also planning to add some random events which would make playing expirience of a single map a bit different every time.


But it all comes down to the avialable time I'll had.
Right now I'll be focusing on basic stuff to make my game athleast playable and then work from there on. This migh just alow me to finally finish my entry in time even thou it won't be best one.


EDIT:
BTW if anyboddy have a suggestion to name my game please do share it as I don't have a clue.
All names that I could think of are already being used

[Super Vegeta]

An alternate option is to create a set of pre-designed maps and randomly select one of those at game start to keep the game from playing the same almost every time.

For greater variety, one could glue a map from a set of pre-designed pieces. Map could be composed of 2x2, 4x4 pieces, whatever. Pieces could either go anywhere, or each piece could have several variants. This way, designing (pieces)*(variants) pieces yields (variants**pieces) maps, and that's a lot. Of course, probably not all variants would fit together, so some kind of matchesWith flag would be needed, but I say it's still worth the benefit.

[WILL]

No updates in a good while. Hows it coming along?

[SilverWarior]

No updates in a good while. Hows it coming along?

Teribly!
To be honest I'm scratching this idea and trying to see if I might have more luck with my second idea I presented in 3rd PGD challenge coments (making a somekinda traffic simulator where the vehicles would be "protagonists").


So far I have reimplemented A* pathfinding algorithm which I now intent to modiffy so that it will take into account more parameters when deciding which path to use (actual distance, esimated travel time for each road section which is based on average time needed for cars to passed certain road section during previous day, etc.). I also plan on making two heuristic levels one will take into account whole road section between intersections and another the actual pathfinding inside each road section to alow overpassing and changing lanes.

Even thou this won't be most optimal I use list of object/classes instead of arrays of records. The reason I decided to do so is ease of upgrading the algorithm with aditional parameters etc. (class inheritance) and with the fact that when using classes I can easily achieve data reusability by using properties. This also gives me ease of implementing mutithreading support.
As you might have read in http://www.pascalgamedevelopment.com...-curve-example I'm thinking of having nicely curved roads of course if that won't require to much time to implement. I can always use grid based aproach which is easy to implement, but won't look as good.

But still I'm afraid I'll run out of time.

[WILL]

Sounds interesting. This would be an interesting project to see come to life.

Good thing you didn't really name the thread then huh? That's okay, sometimes the more creative ideas just don't turn out. The cost of being experimental.

Well if you don't complete the idea, might as well submit something anyway. You at least have the rest of today to get it to a showable state or form of completion.