Prototype Complete...really.

Alright, so I've finished the last few outstanding issues that I had listed in my last post on prototypes progress. The Prototype can now be considered finished from a coding stand point, so all that remains is to work with Microsoft and see if I can't get this past into the next level and to whip my artist a bit in order to get the graphics looking a bit better. (i.e. not the ones I drew :) ).

Play Testing Again...

Ok, so tommorow, on my 24 hours of freedom from work I intend to repair all the outstanding bugs in the Prototype. Now I really do need play testers to find the rest. So here are the requirements:

You must have:

A PC. (The prototype was built for PC to simplify my life).
Shader 3.0 Complient Graphics card.
An XBox 360 Controller.
Preferably a copy of MS GSE, or the XNA redist.
An updated version of DirectX.

OR - failing a desire to obtain, or not actually having the ability to obtain the above, access to my laptop and I. I strongly suspect most of my testers will be in category two.

If you are interested in helping me test the prototype - either post in my Blog, or e-mail me (those who know me should e-mail me).

Thanks.

Nearly Done...For Realz this time!

Ok - so the AI in the game is complete, and because of it I've made a few minor changes to the game play in order to make everything all the more interesting :). Now there are only a few more components I'd like to have in the Prototype before I really can't add anything else to it except better art without a redesign of the game. (Multi-player just isn't feasible the way the code is currently arranged - I have ideas for fixing this - but they will not be in the Prototype).

So - here is the list:

(1) Fix the bugs in the games help system.

(2) Add the difficulty settings and connect them to the Games AI.

(3) Setup the Opening "CutScene" that tells the story of the game.

(4) ????

(5) Profit !!!

After that, I'll drag it around to my happy go lucky play testers and see what M$'s feedback is.

As always, wish me luck.

Still Chugging along...

It must look like I've been working on the AI at this point for pretty much forever, but what can I say - AI is my favourite part of game programming and so it's the place where I will spend the most time :)

I've implemented the idea I was talking about in my last post - there are few coding errors I need to work out ( I am calculating a directional change incorrectly ) however when you take the error into account the AI pretty works exactly as I imagined it. There are few parameters that will need to be tuned - new AI, new balance between difficulty and fun factor after all, but once those are tuned up and ready to go, the prototype's AI will be ready.

With that done, I'll be adding a few more menu features to control some of the other late additions to the games arsenal, finish polishing up the sound (yay for Audacity) and then - maybe, just maybe - the prototype will be finished. Unless of course I decide to add something else, but at some point that has to end right? I've got to show the game off at some point.

Bigger and Badder AI :)

So I just finished working out the main algorithm that will control the games AI behaviour ( this is from the perspective that any new creature I add to the game will use a variation of this algorithm in one form or another ). In order to get the group level behaviour I will need, I worked a way to use a SOM in real time. Basically the creatures will self-organize their overall behavior based on the events they are fed in the game (events having some measure of good or bad of course) - this will produce a final output to which I'll add a little noise in order to get an organized Mob like effect :).

The only creatures this won't work for are the ones that will be working alone - however those ones are special, not fully fleshed out yet, and something you'll have to wait for the full game to see :).

Shader Problem Solved

I solved the problem below :)

More Shader screenshot

This is my shader on drugs.

Wierd Shader Issue...

This is my Shader working normally...

Short update..

Not to worry, I am not dead - however I have put most of my effort this week into getting the hang of my new and 200x more awesome day/night job then I have for working on the game. Now that I've had a couple of days off I can begin polishing the prototype (and it stills needs polish in my opinion). I won't have much new stuff to show off for a while however, as I am working on a larger feature I've just thought up, reworking the story, and keeping tabs on my artist in regards to getting some concrete concept art together. Most of the actual coding is done however - at least for the prototype. Concept art should be availlable next week, which I may or may not show off - depends on what gets done and of course the opinion of my artist.

Warning

So after a few...number theoretical delays involving some prime numbers, a power of two, and the candle stick after midnight - I have now received my NDA with Microsoft. This does not mean that I have the contract, but does mean I need to be more careful about what I put on this site. As such I *may* have to remove some aspects, I may not - I haven't finished reading the NDA yet. When I do, I'll make any changes I need to .

New Models and "Smarter" AI.

Alright - so two new updates my every so increasingly awesome prototype. First - new models, I wanted to cut down on the triangle count and those aliens were just taking up way more triangles then they justified. Second - although you can't see it, I've finished the update on the aliens AI. It's not exactly what I described trying yesterday, rather a hodge podge of a steering algorithm and the skewed probability heuristic I mentioned. It meets all the requirements and gets results I am happy with for the Aliens behavior. Plus, as an added bonus - I now have a simple way of adjusting the games difficulty level by simply adjusting the parameters that control the AI :).

I'm not exactly sure what else I can add to the prototype, the code base doesn't really extend nicely to all too many more of the full games features (since it was going to have to be redesigned for the XBox anyway) so it's hard to say what to add next. I will however keep my possible non-existent readers posted as I clean up the rough edges for the Concept Submission stage.

On Zombies Swarms...

Todays six million dollar question is: How do we model a crowd of Zombies marching towards a target? Our first consideration should be that our Zombies have no brains - so we can't just shove some simple minded implementation of the A* algorithm and expect to come out on top - last I checked brainless creatures don't come equipped with automated GPS - my former employer not withstanding :) .

So how am I going to solve this latest problem? My best candidate for a solution is a kind of random walk - except in this random walk the probability of moving in a given direction is supplied by heuristic. This heuristic should be such that if we take the HIGHEST probability and set it to one and all other to zero that our algorithm will provide the same path as the A* algorithm. Luckily - my game map has no obstructions (yet) - so creating such a Heuristic is easy - simply divide up the probabilty amongst the different directions. As an added touch (inspired by a response I got from XNA's forums) - when the Zombie reaches some fuzzy threshold switch from the random walk heuristic to the A* itself.

I'll be posting on how good the results of this approach are once I get them implemented into the game.

More Improvements...

Alright - although this looks like a major change, it really isn't. All I've done here is merged Microsofts billboard pipeline (changed very slightly to suit my needs ) with a new terrain model to add grass to the game. It doesn't look half bad, but I'd like to come up with ways of improving how it looks. That blue sparkly thing in the middle is the star trek like shield static I was talking about last post. I think it looks pretty good, although it's possible that the static could use a little toning down. Now that I have an artist one board, I'd like to find a way to replace those "MSN" dudes with something more along the lines of the Robo-Zombies they are supposed to be.

I'm really Bad at this Lockdown concept :)

So I wont show any shiny pictures of here today, but I went ahead and made yet another change to the prototype. This time I added cool shader to replace the hack job I was using for the games shield effect. Now that I've got the hang of this whole pixel/vertex shader thing - I can start to make adjustments to games overall look in a more substantial way - graphically, the quality of the prototype is only going to go up :).

Major Change to Prototype...

Well, if you take a glance at the picture, the change should be obvious. I've completed changed the terrain! Why you ask? Well when I originally imagined the game, there was going to be randomly generated terrain with random boundaries. The problem with this is that it simply makes the game untenable - either you'd be lucky enough to get a map you could navigate, or you stood no chance. This was bad for game play, so I never implemented it. However - I left my prototype terrain generator in. Cost - nearly 60 - 100,00 triangles possibly more. The current terrain is a just a big box - it uses less than 20 triangles - allowing for better performance at higher game levels. So what you see is what happens when you remove the useless fat. I think it kinda looks better this way :D

NDA With Microsoft..

So, my contact at Microsoft has mailed out the NDA for the game today. Based on what's in the NDA I may have to remove some of the stuff that's been put on this blog, so if you are looking for posts that used be there and are now gone in the future this is likely why. Once I have the NDA I'll be filling out the Concept Submission information. To help me with this (and the completion of the game) I've conned (asked really nicely :) ) the help out of an Artist. Once I get past concept submission (if?) - then I'll have to "re-engineer" the games code for the move from Prototype to an actual game of respectable quality. That's the hard part of course - and just making sure I've designed the software correctly could take some time - but like everything else, it's hard work because it's cool.

Lockdown.

So it's taken a lot of will power, but I have convinced myself not to make any more changes to the game until people have actually play tested the demo/prototype for me. I've killed all the bugs I know of, balanced the game a bit and made sure all the cool specials that I want in the demo are working correctly. Everything is good in the house of light. Now I need to get my artist on board, and brush on HLSL - which I'll need for the full game. For informational purposes, the prototype demonstrates what would be considered the games "1st level". One alien, standard powers ups, very little AI, and lots of stuff to kill :) . I am not going to say anything more regarding my plans for the other levels, because I pretty much can't right now - but suffice to say that if this works out it's gonna be pretty damned cool!

Another Slow Game Day...

Not much to report on game progress today - with the primary functions I was looking to implement in the Prototype complete, most of the changes I've made are small optimizations. I've moved a few matrix mults. to the GPU, removed some extraneous geometry, and reduced the load put on the Garbage Collector. The game should run a bit smoother now, but there are still quite a few more places that could use that optimizing touch :)

Progress on the Publishing front is also slow - I've heard back from Microsoft - they're putting together an NDA for me while we wait for the Account Manager to return with some feedback. If they are anywhere near as busy as I think they are, then I may not hear back from them for a few days.

Prototype Playtesters

Do you wanna be the envy of your peers? That manly man that women simply can't resist! Yes - we thought so - that's why we are offering you this amazing opportunity - become a play tester for Robotic Zen (the code name for Gene's Game Company ) test the awesomest game to hit the arcade for the brilliant salary that is exactly equal to the number of privileged reference frames in this universe! Yes all this can be yours - sign up today!

Partcle Beams for the Win!

The last special effect needed to complete the set of features I wanted to include in my prototype has now been integrated - the all mighty particle beam! Revel in it's pen-ultimate coolness... ok you can stop now.

Now this becomes a waiting game - hopefully I'll get some feed back from Microsoft soon, they e-mailed me saying that would get back as soon as possible. Wish me luck.