The theme this year was “Overload” – our game concept was basically this: the major tech corporations are all working together to take over the world, you player a “hacker” and try to take down their servers by overloading them.  You do this by moving around inside the “server”, while a bunch of “minions” follow you around and do the heavy lifting (a lot like Overlord or Pikmin)  We called it DDoS (Distributed Denial of Service, a technique used to do pretty much just that).  Menu’s and such, while they turned out very nice, and were a lot of work, weren’t really impressive to any non-programmers, and their development sorta spanned the entire development period.  We used XNA with C#, so we could  develop more easily for the PC and XBOX, but we wrote our own engine from scratch – graphics management, input handling, collision handling, etc.  The engine is really, really solid – once we get a chance to clean it up a bit, maybe make some small changes, I’d really love to pull it out into it’s own project, so we can later use it as the base for other games.

The menu’s were primarily Brian’s work, and while they’re quite impressive, I don’t know enough about them to discuss them at length.  The immediate game-play, how rooms are handled, how connections between rooms are handled, etc. was primarily Josh’s work, so again, I don’t really know enough to discuss at length.  My work lies primarily with the gambit-based AI system.  It was based pretty strongly  on the Final Fantasy 12 Gambits, or the tactics system from Dragon Age.  Basically, all AI controlled units get a set of “Gambits”, each of which has a condition, and an action to be taken if that condition is fulfilled.  These could be things like “If an enemy is withing 200 pixels, shoot at them”.  The object has several of these, which are checked in order, with the higher  ones taking priority.  So if the previous example was in the second slot, and in the first slot was a Gambit that read: “If a nearby ally has less than 25% health, heal them”, then any time a badly hurt ally was nearby, the unit would concentrate on healing them, but if all the allies were  okay, then it would attack any enemies  that came within 200 pixels.  With a large enough set of gambits, this can become a very powerful and flexible system for AI.  Since I’ve set it up such that any condition can be paired with any action, we’ve got a fair-sized list even with the 2-3 actions we’ve currently got.  We had planned on letting the player customize the gambit list for his little AI controlled minions, but didn’t have time.

Development of our game actually went really, really smoothly through Friday and Saturday.  It wasn’t until the last freaking minute on Sunday that we found all these nasty bugs with the collision engine, and didn’t have time to fix them… competitive fail, but I think we learned a lot even from that.  I learned a lot about programming, but I also learned quite a bit about the team and project management.  First off, I think for future projects we’ll stick with Tortoise SVN, cause Anhk is really buggy xP  Second, come the Spring competition, we’re going to spend more time sleeping   I only slept for 2 hours, Sunday morning.  Thing  is, if you sleep for even 6 hours a night, you give up a quarter of your development time, which we really can’t afford to do.  That said, I think if we managed, say, four hours a night, or three 2-hour naps each day at different times (so we’ve always got at least one person working), that the lost time  would be made up for by the increase productivity from being better rested.

That’s all I can think of to say right now, once we get a more workable demo, maybe I’ll put a link to it here, if the guys don’t mind.

Anyways, when you’re given so short a time to work on something like that, every possible minute is spent working.  All but two people on my team went home the first night, and all but three the second, I was in both groups that stayed.  I didn’t sleep at all the first night, although I slept from 3-6AM Sunday morning, and about an hour during the drive back to the dorms.  I didn’t get to do as much actual programming as I would have liked, due to my lack of experience with XNA, although I’m substantially more comfortable with it now than I was when I got there.  That said, I did help with basically all of the math, both for the platformer physics and the mastermind mini-game.  We developed the mastermind mini-game first, but we only had it set up to use text output, simple debugging text to make sure it worked, so I later re-wrote a portion of it to give it some semblance of a GUI.  Only code I really wrote, but whatever works

I picked up quite a lot, both in how to do things (mostly design stuff, to be honest) and how not to do things (like, this is how not to lay out a platformer engine within 48 hours… in fact, don’t try to write even a simple physics/collision engine from scratch in 48 hours ).  I’m going to, I think, spend a fair amount of time tomorrow starting to rewrite much of it, in an attempt to make it cleaner, and because if I can pull off a platform engine of any kind in XNA I have a potential XBLA game I might be able to put on the XBL marketplace….  Anyways, in the list of things I learned how to do (or how to do better): Agile development methodology, one way to lay out a state-machine based game (I like my way better though ), a lot of stuff in/about C# and XNA and and work while tired  In the list of things I learned how not to do: Write a collision/physics engine, store level data, sleep (well, rather, learned how to not sleep ), and I learned a lot about how sloppy code can get when you’re so limited on time  If such things were decided by the quality of the code I’ve written, I would be burning for eternity for some of the ugly hacks and work-arounds made over the course of the weekend

I also learned about an awesome (and incredibly geeky/nerdy) card game called “Munchkin”, and made a lot of progress in FFVI in my spare time  I learned that there really is no good place to be in a building full of guys after a large taco-bell meal  I also learned a lot about sleep-deprivation (yes, in an attempt to stay awake, I decided to read articles on sleep-deprivation ).  I actually, to a very small extent, experienced a lot of the side effects from minor sleep deprivation, things like micro-sleeps (where your body shuts down AKA sleeps for about 10-30 seconds before waking back up), some shaking and actually a noticeable temperature change when really tired, go figure  Anyways, in my reading I also found something about a condition that I’ve decided is probably the worst possible way to die, and basically what it is, is you slowly lose the ability to sleep, causing you to lose more and more sleep, which actually drives the person insane, before eventually just killing them from exhaustion….  Off that depressing note, once I worked through the initial exhaustion, not counting the micro-sleeps I was basically perfectly fine the rest of the time, even the second night, it wasn’t bad, I slept more because the rest of the team was and I couldn’t do anything until they got up than because I felt tired.  I actually don’t feel tired now, although I don’t really feel awake either, it’s sorta like I simply can’t tell, something that will probably be remedied in the morning

LINK

Anyways, I won’t bore anyone further:

LINK

DOWNLOAD

That being said, as of a few minutes ago, there was a new Tower of Azari release (0.306).  Progress since the 0.301 release is staggering.  About the only feature from the pre-rewrite releases that 0.306 is missing is saving and loading.  Magic is back and far more interesting, and everything just generally looks better and runs smoother.

I won’t bore you with them here, but there is extensive information about the last few releases in the discussion topic here.

Download

I’m going to look into only requiring approval from people who haven’t posted before, hopefully that’ll make it easier for anyone who wants to comment regularly, while still blocking most of the spam…

Note: Version 0.301 is a tech-demo, and is far from being fully playable. It has a working menu, map generation, movement, doors, message log and some other things, but there’s no monster to kill, no loot to collect, and no goals to achieve. Version 0.30 is mostly functional, if a little buggy, so if you’re wanting to try Tower of Azari for the first time try that one first. That being said, if you’re interested in seeing where I’m taking it with the re-write, check out the newer version.

Still interested?  Here’s the link!

Internal Changes:

• New item type, “TYPE_MAGIC”
• New item attribute, “item_spell”

Visual Changes:

• Fixed gold being drawn only when out of sight

Gameplay Changes:

• Magic is in!  Press ‘C’ to cast a spell from memory (if you know any)
• Use a wand (W-*) or scroll (S-*) item to cast it’s spell

Internal Changes

• New item type, “TYPE_AMMO”
• New item attribute, “ammo_type”
• Class information is now loaded from a file for character creation.

Visual Changes

• Made items brown on the map, to differentiate them from enemies.
• Fixed items sometimes not being drawn.
• Item purchase screen updated to allow the purchase of multiples.

Gameplay Changes

• Ranged weapons are in, they can be used as melee weapons, or fire projectiles.  Press ‘f’ to fire.
• Ammo will stack when picked up, purchased or fired.
• Enemies will not block a shot, so you can shoot past one enemy to hit another.
• Fixed doors being placed on the right and bottom outer walls.
• Look function now lets you pick a square to look at.

Editor Changes

• Item editor and item.raw format updated for ammo_type
• Icons added for editors