Log in

No account? Create an account
26 July 2007 @ 10:13 pm
WEEK 2 - Pre-production  
Today was a day of semi-organised chaos. It was Open Day at the AIE, and that meant people wandering in and out of the lab all day, harried lecturers running about, and admin staff extolling our virtues as we sat and attempted to look like we actually knew what we were doing. I'm not sure if we actually managed to fool them ^_-

First thing in the morning, we got bad news - apparently there is already a student-made game called "Lobot". I looked it up on youtube and there's a trailer for it. It seems to be a pretty standard platformer where you play a cat that shreds things. What that has to do with the name 'Lobot' is anyone's guess. But it does mean that we're going to need a new name! Suggestions are welcome, because I have no clues ^_^;


Today Ben continued work on his test framework. He managed to get the Aporkalypse world model loading, and then set it up so that he could create and drop cubes into the world while the program was running, and watch them bounce around. Actually seeing progress was very cool. It felt like we finally broke through the 'we have no idea how to even start this' barrier.

Meanwhile, Wayne got the wireframe of the physics world model appearing, so you could actually walk around and see how close it was to the graphical model.

Peter started work on the NPC AI, creating a skeleton rule system for the behaviour of different kinds of bots.

Shane spent the day wading through the animation loading and pipelining code in Aporkalypse. He came to much the same conclusion I did - it's messy, ugly, great chunks of it are from earlier experiments and are now redundant, and it would probably be easier to re-write the entire thing from scratch than try and clean it up. Daniel, however, disagrees, pointing out that at the very least, it does appear to work relatively reliably (it's just a shame the same can't be said for the rest of the Aporkalypse code). We're still undecided, but Shane will probably have the final say, since he and I are the ones that will be working on the Lobot code, and he's the more experienced programmer of the two of us.

On the other hand, the ongoing NovadeX/Ageia PhysX/no physics engine debate was finally decided. We're going with Ageia PhysX, since we managed to find a plugin for it that says it works with Max 9. That'll at least remove one step of the tangled, time-consuming exporting process. I guess we'll deal with the 'need to install program before the game can run' issue as it comes up.

Ben installed Ageia on his computer, and then went through all of the demos, to see what kind of upgrades the engine had had since it's NovadeX days.

There was an awesome stretchy frog that you could pick up by various bodyparts and throw around, which demonstrated some seriously nice deformable soft-body physics. There was also a cloth demo, which was really, really realistic. It was apparently created by joining huge numbers of very tiny particles together with springs. Probably too computationally-expensive for us to use in our game, but a very interesting demo of what can be done with the engine.

Ageia also seems to have some kind of in-built sensing/reaction scripting. One of the demos had cubes in it that had a sensor just in front of them, so that if they were moving forwards and about to crash into something, they'd turn to the left to avoid it. It was pretty simple, as far as scripting goes, but qutie effective.

We probably won't be using that particular feature, though. We need to keep all of the AI-related stuff in the actual AI code section, rather than scattered throughout the physics middleware ^_^;

Anyway, despite getting Ageia working on its own, Ben couldn't get the plugin working with Max 9 at all. After some investigation, his best guess was that the plugin was for an earlier version of Max, and that we'd have to try and look for a newer one tomorrow.

To add to the software frustrations, Shane-the-graphics-lecturer-not-the-programmer couldn't get Renderware working on the dedicated Max 8 machine, either.

Hopefully tomorrow will have fewer technical glitches.

Game Design

Geoff and I spent most of the day with the design team. In his morning speech, Paul informed the class that tomorrow the design team would be giving a presentation and walkthrough of the entire game in front of everyone. This was news to us, and rather stress-inducing.

But, despite still not having an exact idea of what we were supposed to be producing, we soldiered on. We started creating timelines for each level, showing the order of locations within the level, the goals for each subsection, the skills that the player would aquire there, and the type of puzzles that would appear in the subsection. We got the first level done and started working on the second level.

As we were working things out, we got into a discussion about whether we should have a 'levelling' system, and if so, how it should work. Levels are useful in one way, as they can give the user something to aim for, as well as helping them feel like they're actually achieving something and progressing. We tossed around the idea of NPC robots having levels, and Lobot being unable to deactivate anything of a higher level than he was.

The problem with that idea was the question of what to do if Belfry is set upon by a robot too high for Lobot to deactivate? Neither Belfry nor Lobot have any way of getting rid of it, which means the situation results in deadlock. Belfry can't move with the shield on, but he can't turn the shield off, or he dies. Guess that's another thing to try and iron out tomorrow.


We did some more experimenting this morning, trying to get multi-texturing to work with Max 8, but with no success. This may have been because it was mostly Paul and the programming students trying it, with advice from some of the graphics students floating nearby. It's quite possible someone more experienced than us at Max (ie, Eve) would have more luck, but she was so busy with Open Day happenings that we didn't manage to grab her all day. Guess we'll try again tomorrow.

Today the artists have been refining their designs, and drawing even more detailed pictures. A couple of new Belfry designs went up that I liked, complete with crazy hair. Still no final decision on Belfry or Lobot, though. I guess we're still only halfway through preproduction, though, so that's not entirely unexpected.

So, that was it for today. More tomorrow!
(Anonymous) on July 28th, 2007 06:33 pm (UTC)
Names o_o;
TN here =D

Thanks for all these weekly updates Michelle! These certainly give a fantastic overview and insight that kills my curiosity, rather than the cat in me ;P

Names names names... Hmmm. If ya haven't nailed down any new names yet - Spontaneous thought time~

BBB? Buddy Bot Belfry? (or some variation of that?)
Fix Bit? FixBits?
- BotFix?
- CodeFix?
Breakpoint? (Relating to coding and what Lobot might be doing to deactivate other robots?)
- BreakByte? ByteBreak?
....yeah XD
If anymore come to mind, and still haven't read confirmation on names ;P I'll letcha know.

Keep things up =D You're gonna do awesome!
Thanks again! =D catch up with ya soon.
aie_michelle on August 1st, 2007 01:31 pm (UTC)
Re: Names o_o;
Hehehehe - I'm glad you're getting something out of them, at least ^_^

Names are hard, and we still haven't chosen one yet. Thankyou for the suggestions - I've added them to my list =) If you think of any more, send them my way, ok!
(Anonymous) on July 31st, 2007 01:19 pm (UTC)
Terry (not anon)
Great write up Michelle. Keep em coming Ah you spotted that problem with Belfry's shield a day before we discussed it :).
Hi TN!

cheers Terry
aie_michelle on August 1st, 2007 01:28 pm (UTC)
Re: Terry (not anon)
Yes! I just didn't have any answers to it until Geoff started ideas my way *grins*

Glad the writeups are useful and/or interesting! =)