Ravuya's Articles!

What went right

1. Good base -- I had good technology already built up by the time I started the project. A good portion of CSRPG2's structure was carried over from CSRPG1, and what didn't work (i.e. the maps, the animation system) were tossed out and replaced with code from my action games.


2. Good editor -- I went back and modified the editor many times over the course of the project to add a fill mode, a minimap and a palette of available tiles. I also added an entity editor (a first for any game editor that I have ever built) which greatly unified and sped up development. Well worth the time I spent on it.


3. Fan base -- People wanted me to finish the game. People wanted to run around and play, they wanted to know what I was doing with Dr. Questy. When I thought of them, design decisions just "clicked" and it was easy to make decisions that would please them and me. I felt like I was truly building a universe that would be playable and enjoyable, rather than one that tried to be too serious and tackle serious life issues (I'm looking at you, Final Fantasy VIII).


4. Humour -- From getting piss drunk to watching the sea monster beat things apart, CSRPG2 kept you laughing from start to finish. People still tell me the ending is one of the finest they've ever seen, and I agree. Humour makes sure you keep having fun, looking out for that last little secret, which brings me to my last point...


5. Focus -- I was extremely focused on the goal of CSRPG2 -- Exploration. Unfortunately, a number of places (like RavLand, the happiest place on the lower 72 quadrants of the planet Earth) were shelved to produce the game in time for my final exams, but there are still lots of places to see, things to do and people to touch. Plus, I only used like 1/6 of the available map for the game. There's plenty of room for more content in the future, if somebody wants to add it.

What went wrong

1. Art -- I don't have any. While using animated stick people made the game move a lot easier (and be a lot funnier) they also looked absolutely terrible and probably detracted from the game. I need to get an artist, stat! More flash probably would have gotten CSRPG2 featured on Game Hippo as well as other sites that would have moved my downloads into the dozens of thousands (current download total: ~1500, about 300 of those for Macintosh).


2. Windows port -- I didn't take much time on it, and as such a number of bugs regarding Windows' handling of floating-point values slipped through. There are a number of collision bugs here and there that don't exist on the other ports. Unfortunately the Windows port was my largest user-base and it probably reflected poorly on me.


3. No Scripting -- I waffled on and off about it for months even before the project entered development and I am still completely surprised I didn't just sit the fuck down and code some of this shit. I spent a lot of time hacking up some things that should have been done with scripts (the Sea Monster) and I probably would have been able to put together some nice scripted jokes if I had just taken the time to code the fucker. Instead, I hacked until I got a good game.


4. Combat AI -- While I feel the combat system is one of the more unique components of the game (and will likely be improved upon for future RPGs), the enemy AI is absolutely abysmal. Even when divided by a river, the enemy will not use projectile weapons or any sort of magic. It would have been a far more challenging game if the enemies had in fact used better tactics, but this would have required further hacks on the already terrible NPC movement code.


5. Interface/Documentation -- This was my Achilles heel. After release, I found myself constantly being asked "so, how do I play this" or "how do I open doors?". I added a training mission but that really wasn't enough to solve the problems. In addition, getting close to a door, then selecting the key from a separate inventory subscreen didn't make a lot of sense. I will be going for a less modal interface next time around, potentially with annoying Xbox game-style "hot tips" when you approach new objects. Offering a readme, or some online documentation, would probably also have helped here.