For a long time now I've been keeping things stupid-simple when it came to on-disk layout of my source tree. With a large set of files this started to get a bit unruly. The real reason for doing this was to make my life simple when compiling for windows (not having to deal with makefiles). Well, those days are over. I did some major reorganizing. That was the easy part. The more difficult part came when I had to update the Windows build to "do the right thing" with this new structure.

This meant dealing with Makefiles -- something I don't really have much skill with at all. Long story short, I finally worked my way through it (with some help, of course), and things are building just fine now. I haven't tried anything for a linux build, but that's low priority right now and I likely won't look at that until much later (sorry Linux lovers).

With a working build I can get back into other things that don't involve figuring out how to compile what, when. For that, I'm grateful. I would make a horrible build engineer.

I have been on a very long hiatus from Neverfall due to other work. I am hoping that soon I can start getting back into the groove. Likely it's going to take some time to ramp back up again, though. I should find something relatively small to ease my way in with. That said, I'm not really sure what that would be, exactly.

My plan was to hopefully finish Neverfall by this Christmas, but I think that goal is unattainable at this point. Maybe next Christmas? Here's to hoping! I've been working on this project for much too long. It should be complete by now!

Looking back at my old tilesets it was readily evident that the first one I completed needed to be redone to match the rest that followed. As I progressed through each tileset the amount of detail increased as I actually gained more skill in pixel-art--at least when it came to tiles. The first set didn't have the same feel as the rest at all, so it got a complete facelift. The overall structure stayed the same, it just seems more "textured" now.

After throwing that tileset together I began working on the first NPC type. To be honest, I wasn't really prepared for having to do even more art. I have a basic animation for movement and idling that works out quite well. In putting these together, however, I realized that generating both "happy" and "sad" versions of every single frame of every single sprite would be a lot of work. I'm opting for, instead of doing that, to have an indicator that floats near each NPC depicting whether they are sad or not. This way, it's easy to tell that something needs to be done. Once they are made happy, no more indicator. I think it might work out quite well, but I need to throw together a prototype version of it to play around with in order to be sure.

On top of that, creating a rudimentary AI for these little creatures is going to be.. "interesting" for lack of a better term. So far each NPC type has the basic ability to walk left and right, while colliding with the world around them. Some behaviors they don't have are ones like jumping over an obstacle in their way. I definitely do not want them moving between maps on their own unless carried by the player. I'm hoping I can get by with making them as zombie-like as possible while still being convincing.

One step at a time. Next up is showing a sad-badge on top of sad NPCs.

Finally "connected" one end of the game world to another. By that I mean: beginning with the starting area, I have created one of the first loops where you end up back in that same starting area without completely backtracking. This actually paves the way for the first big chunk of backtracking in the game; at least, backtracking how I envision it working in an almost completely untested game world. Once I start testing it out, more likely than not I'm going to have to start tweaking things here and there. So, this basically means that I will be taking it easy on world-building for a while. Now that I have this first loop set up, I need to start populating the world and ensure that what I have setup is workable. That means I'll be looking at the following stuff:

• Fixing and cleaning up tilesets that need it
• Tweaking maps with the new cleaned-up tiles
• Placing objects! Powerups, NPCs, save points!
• Testing and bugfixing

To be honest, it feels pretty good to know I will be taking a break from world-building for a little while. It's not torture, but after focusing on it for so long a break sounds really nice. Playing with object placement is also going to force me into areas of the codebase that I haven't touched in a while. This will be good, since there's still a lot of functionality and entire objects that I know I will need yet are still missing.

I've generated enough of the green caves tileset to start building maps with it. There is still an incredible amount of work left in terms of world building. At first blush it didn't seem like it would take quite as long as it is. Level building itself simply takes a long time. I'm beginning to wonder if there might be some easier way to speed up the process, but it seems unlikely. In a tile-driven world, you really do need to place all the tiles piece by piece. Maybe I just need to implement a various tile-sized bucket fill?

Other than world building, I updated the Neverfall page with some new screenshots that use the current tilesets.

A few bug fixes this time around. First, an easy one: map background image coordinates were saved out incorrectly. Thankfully, it hasn't affected any of the previously saved maps. That's good! Turns out I was saving the texture height as the texture width. All background textures so far have been square--so I've been lucky! I noticed when I tried using a new background I hadn't used before. That one was a freebie.

The control mechanics for walljumping has bugged me for a while now. When you jump into a wall and then push away, the user "sticks" to the wall surface for a moment to allow you to execute a wall jump. This didn't always trigger correctly (or sometimes triggered when it shouldn't). This has been updated so it's less optimistic, allowing the player to maneuver through really tight spots without always clinging to walls. In addition to that, it wasn't always obvious (to the player) why their character was sticking to the wall. It seems like a bug. You'll now have a visual indication of when your player is "ready" to do a walljump. I think, overall, the walljumping is going to be the number one control issue with the game. I'll need to give it some more attention to weed out all the little quirks.

Outside of coding land, world completion rate is at around 33% for maps. That's good! I hit a roadblock, however, in that I'm now at an area of the world that doesn't have a full tileset. Due to that, it's time to put my artists hat back on. I've been tiling and extending the previous set I posted a while back...

It's been some time since I seriously tinkered with the codebase. This time around I tackled a few issues. The first was the saving and loading of the world objects. There was a bunch of missing gaps that I'm surprised I didn't notice until just now. The worst offender was that certain object types simply were not writing out to the file, and, if encountered somewhere in the middle of the list of all the objects would cut off writing of the file. Very bad. Needless to say, thats been cleaned up. On top of that, I finally added in some versioning to each object that attempts to write itself to the master list. Moving forward this will be really important so I can enable writing of a new version while still reading in an old version, thus preventing the necessity of having to obliterate the master object list and recreate it from scratch. Versioning is good! It's something I don't do enough of.

In the realm of adding functionality as opposed to fixing bugs: I enabled mp3 support in the game engine. This is something that's been around in my framework forever--but I simply never added wrappers for it in the game's engine. It's there now, and I'm experimenting with different types of background music (not my own, of course), to see what style fits the best. I really need to put my musician hat on and start generating some new music. Either that, or find someone else who has a better musician hat than I do!

Lastly, I revisited the save game browser. The first incarnation was a usability nightmare. Originally it worked like the following:

• First item was "exit" (backs you out of the browser)
• Next item is the "oldest" save created
• Followed by each newer consecutive save creation date
• Last item was the "New" option, to create a new save

Now, it's a lot more usable:

• "Exit" is still first
• "New" save option is the second item
• Followed by the most recent save created
• Followed by each "next oldest" save creation date

Note that one can save a new game over an old game. These don't move up or down in the list. Basically the new usage pattern is a lot more natural. If you want a new save, you simply move to the second item in the list. If you browse the list, that most recently created save will be the next option. It's more natural to want to load a more recent save over an older save, therefore the most recent options should appear at the top of the list. Much better!


In addition to those changes, the save game rundown contains your current total playtime, total number of saves for the given game, and a datestamp of the last save. Pretty spiffy!



This gem is a remnant of the old save browser that has yet to be fixed. The thumbnail image has yet to be updated to reflect the new style of the player sprite. It's really interesting to see how the entire game style has changed over time. I remember the days when there was no visual direction to be spoken of at all.

That "back" arrow really needs to go, too!

I've been putting together more maps, but progress has been slow. To steer myself away from getting burnt out I played around a bit with the in-game menu. This is the menu that you can bring up to let you know how much of the population you have helped, what powerups you currently have in your possession, as well as how long you've been playing the game. It slides in and out of the top of the screen when summoned.

It's pretty much the same information as before--it's just presented in a much nicer fashion now. Some of the contents may need to be altered as I have yet to decide whether or not certain races are going to make the cut. At some point I will start populating the game world with inhabitants. At that point I'll see how repetitive things will get and whether or not some of the races will be a requirement or not.

I've been silent for quite some time now--but busy! After finishing creation of importable versions of the map I set to work! Initially, it was actually really fun seeing the game world come to life. Being able to traverse through areas seamlessly is very rewarding--especially when I know that these areas will likely make it to the final version of the game. It's a bit different when you're playing around in a demo world that you know will be trashed at some point. There's just a different feel to playing around in the world!

So far, map completion rate is at about 30%. This includes laying out the map, filling it in with background and foreground layers, and other "decals" (for lack of a better term). It does not include things like placement of NPCs or objects. Actually filling in the game world with inhabitants, obstacles, and powerups will come later. Of course, I've already planned out the general location of all of these. It's all subject to change slightly, as you might expect.

At this point, though, map creation is becoming slightly tedious. I've been at it for a while (obviously with breaks in between since it's been silent for close to a month). I can definitely say that by the time I reach the last few maps of a given area I have started to grow tired of its tileset. Needless to say, I'll be buffering areas that use the same tileset with areas that use a different one.

I'd love to post a nice capture of the completed areas contrasted with the entire world map, but I don't want to spoil the structure of the world just yet.

World building has taken off to a slow start. I've spent a lot of time compiling importable versions of the entire world map. Unfortunately, due to how my image->map importer works, they cannot be exact size (power of two requirements). On top of that, when filling in the initial base-contents of each map I wanted to them laid out spatially so I can see how they all connect to one another. At the end of the day I created roughly three different representations of the entire world map, all useful under specific circumstances.




In doing map imports and general editing work I realized that there were a bunch of overhauls that needed to be added to the editor. This includes things like: not reinitializing the editor every time you switch between edit and play mode, remembering previously saved to locations, saving opened files without bringing up a dialog, "save as" ability, etc.

Aside from world creation I also implemented triggers for stopping the camera at certain points while moving through a map. This allows me to give off the impression that there is no more map to be explored while in actuality there is data and tiles in a certain position. This allows me to create oddly shaped maps where, when the player walks up to a wall, the camera is stopped instead of showing a whole bunch of filled in tiles.