The one year mark of this project would've also marked the 10 years that I've been working from the same code base.
My Code base's 10th b-day
So 10.5 years ago now, I started writing the code that I'm using now. Back then I was looking at getting some classes to handle 2d animations and generally try to keep things organized such that animation data, graphical data, and game related data for characters were all kept separate and manageable. For the most part, the animation class has changed very little over the years (changed from using arrays to vectors awhile back and just recently a change to support different animation types better) while the other two have evolved from one kind of organized chaos to another, spilling into additional classes and sub-classes.
Switching to SFML
One change that was long overdue I made last September and that was switching from DirectX 7 to SFML 2.1. If I remember right, DX7 was obsolete when I first started with it. But it so happened that I was able to wrap my head around the code I found for it well enough back then and all I really needed was to just get an image to the screen. And it did that well enough for my needs for quite awhile so I never bothered looking for anything else. Finally though I found with this RPG project I'm blitting enough things to the screen that it was really slowing things down on my target computer (which is also quite old). After a few attempts at trying to make things more efficient and looking around for options, finding out how much things would improve if I was taking advantage of hardware acceleration.
I went with SFML because it was the easiest to get working. Or at least, it was the first option that I got working. I found both SFML and SDL streight forward enough to get an image to the screen but I was unable to figure out how to make use of hardware acceleration with SDL. I would've experimented with an updated version of DirectX or XNA but I was limited by my dev machine. In the end, SFML was the easiest to install and see an improvement with.
Most of the past 6 months has been messing around with movement. I already had a system for using A* that is working alright for me but there was other stuff to do.
Walking through forests.
Since my last post I've changed from using a keyboard to move the player to using the mouse. I did this because I like the fairly thick forests that I have in the game world (see video in previous post) and I want to have it so that the player can move between the trees reasonably well. It was just way to easy to get caught on a tree when moving through a forest using the keyboard. Some quick experiments with mouse movement showed that being able to travel at angles other than 45 and 90 degrees really helped. The path finding stuff that I had didn't have the resolution to navigate between the two or three trees that could be found on a tile so it took awhile to get a sort of steering behavior and path finding combination working to make travel through forests reasonable. It's still quite possible to get a bit caught on trees but I'm thinking trying to move through a forest doesn't have to mean flawlessly traveling to your destination without the need for changes in course.
Fast travel element added.
Before that got done though I was sidetracked by an idea I had for a fast travel sort of thing. I'm not sure why I started thinking of it but I figured if I needed such a feature I'd want to avoid having it like you teleport from one place to another. As it is, I've occasionally generated 1024*1024 or 2048*2048 maps to play with and walking around a world that large does get a little tedious. I ended up laying down roads from the center of the map to random locations using the A* stuff that I had to try and find a path such that roads are 1 tile wide and are laid down so that changes in elevation from tile to tile are minimal. Now, when a player is on a road and does a mouse-key click combination, a flood fill sort of thing is done down the road tiles to the next point an intersection is found. A path is then set to that destination and the player is moved along the road at an accelerated speed. This way, you can follow a road quickly without knowing where it's going and you can always interrupt the travel to go off road.
I don't really know if roads are going to play any part in whatever I end up with for a game but it was an amusing diversion and I can use the feature later if I choose..
Future development plans
When I started this project I was thinking the main character would be trying to learn magic from the ground up. So I wanted to be able to show hand gestures without having to create sprites for every imaginable combination. Since my main character is basically a couple minor evolutions away from a rectangle body with a circle head, I figured why not a couple more circles for hands that are sort of disembodied. Not worrying about drawing arms, I should be able to float these hands around in whatever pattern I need them to and hopefully get a versatile way to animate interacting in different ways.
So far I've just worked on having hands sway by the player's side as he walks. Still a fair bit more to do here that will take my focus for the next while.
Combat (lack there of and some planning)
Probably the main thing holding me back from adding combat into the game is that it seems like I don't have a reason for the player to fight. And that's pretty much because I'm not yet that sure what it is that the point of the game is. The slimes that I've added so far aren't aggressive to the player and I can't really think of a reason for them to be unless I embrace the notion of making a hack and slash sort of game. There's nothing wrong with that but it doesn't really seem to go with the other features that I've developed so far.
But awhile ago, somebody posted, I think it was on their journal, that they were switching from a mouse driven movement system to a keyboard one because they found that it provided a better overall experience when it came to combat. I'd link to that post but I can't seem to find it. Whoever you are, thanks because it made me realize that I'll be facing this problem too. In my case though, I'm not willing to abandon mouse movement because I want to keep the whole moving through forests element that I like. After thinking on how to do combat for awhile, I'm leaning heavily towards something where the PC is kept facing the mouse cursor and the player would press a key or mouse button to execute a melee attack. The attack will have a range that I guess could be described like a radar sweep of various widths and lengths depending on the type of attack. Anything within that sweep will be considered hit. I'm hoping that this idea will keep the pace of the action that I currently have.