Upcoming Events
Unite 2010
11/10 - 11/12 @ Montréal, Canada

GDC China
12/5 - 12/7 @ Shanghai, China

Asia Game Show 2010
12/24 - 12/27  

GDC 2011
2/28 - 3/4 @ San Francisco, CA

More events...
Quick Stats
49 people currently visiting GDNet.
2406 articles in the reference section.

Help us fight cancer!
Join SETI Team GDNet!
Link to us Events 4 Gamers
Intel sponsors gamedev.net search:

Small Scale Development, Grand Scale Ideas


9:12 PM 11/25/2004

With most of us having no accessible family nearby, three of the four members of Codus Vivendi had Thanksgiving dinner. Our initial plan was to get a pizza, but our normal pizza place didn't answer their phone. Next, we tried Burger King, and then McDonalds, only to find that they were both closed. In retrospect, it makes sense that Fast Food joints would be closed on the only holiday on the calendar set aside exclusively for food. I presume that either the cartoon network marathon or the excessive hunger temporarily disabled the logic sector of our brains.

On the way back from the last fast food place, I remembered I had all sorts of food at my place that I hadn't used since I was never there anymore. We dropped by and I picked up all the things I told myself I was going to cook a few weeks ago when I went shopping. We got back to JeffH's and put everything together, and made a full meal.

For the first time since my last visit with my relatives, we had dinner; complete with a conversation over a real meal on a plate at a table, instead of Burger King from a bag in front of the T.V. JeffH said that "We're like a family now", in a really stupid tone of voice. I just looked at him.

7:24 PM 11/26/2004

In light of recent events, we've found that the game models may need to be modified to some extent. Peter has the game models responding to NovodeX but he believes that the orientation of the joints on the model is what is keeping his death animations from being entirely accurate. His system is set up so that the parent joints point to the child joints with their y-axes. It doesn't look like the model adheres to that rule completely, and the .mb file of the model shows that. We still need to talk to our artist since our artistic abilities are limited. Unfortunately, we won't be able to speak with him until Monday. I decided that Peter's time would be best spent on something that he can complete now, without waiting or doing any heavy trial and error. His next scheduled task is making the Audio Wrapper, so I'm assigning that to him in the interim.

8:14 PM 11/28/2004

The AI is acting funny. For some reason when an entity is going through its list of scripted objectives, it ticks through its movement instead of moving fluidly. I'm using the Hi-Res Timestamped Event Logger JeffM wrote, and still am not certain why it's happening, but plan to figure it out soon.

Friday, Peter started work on the Sound Wrapper. Today (Sunday) he is almost finished. As it turns out Audio isn't as heavy as I had scheduled it to be. Once again, my scheduling pessimism has saved the day = P … and yeah, Peter codes fast too.

I also received an email back from Rory Young, telling me that realigning the models wouldn't be much work. I didn't believe it would be hard to begin with, and with the way that Rory cranks out game models, it's barely an issue.

JeffH has been working on combat animations for some time. I'm making more room in the schedule for it because of the following reasons:

  1. Good Animations are core to the game, if they're ever bad, no amount of AI, Particle Effects, or Ladder Rankings can make up for that.
  2. JeffH will make up for lost time once the new system is set up.

But you might ask, "What is this new system?"  In a nutshell, our animation system is based on entire full-body animations. When one stops and another starts we blend between them. That way when you go from running to standing still, the character doesn't just snap from one frame to another, he blends.

Right now, when you have your characters running, and casting, and stopping and shooting again the blending gets complicated. Currently, if you hit the attack button while running, your character looks like he's stepping over a mud puddle and under a limbo bar at the same time. Scary Stuff.

Our solution is splitting each Entity model into a top and bottom half, so that we can isolate the running and attacking animations to allow the player to do both at the same time, but smoothly. (See Figure below)

Hopefully it won't be too hard to split the models into two halves and re-export their animations.

8:56 PM 11/30/2004

JeffH and I talked to Rory today about the models being fixed. After we got the basics of the model changes we needed down, I sat back down. Somehow I think he got the idea that he would have to redo the model from scratch – which he wouldn't.  By the end of the day he seemed visibly upset when I spoke to him on our way out. He isn't going to do anything until tomorrow though, because we still have something else to try before splitting the models, which is a last resort. I sent him an email a little while ago to him for general damage control, and to clarify what we needed and what we wanted to finish. All of us appreciate his work and we want it to look the way it was meant to be. I do not want to use a version of his model that doesn't export to RenderWare correctly when all we need to do is get a work around for the subanimations.

The model is _almost_ there. It looks fine in maya, but something gets lost in the translation. For some reason, the joints aren't all exported to RW, but most are. JeffH can finally separate animations on the basis of their parts, and he needs to be able to identify specific joints so that he can turn of some of them off and leave others on for specific animations. If he gets all the joints we might not need to split the model up like we said.

Peter is back onto NovodeX. It seems that going back to the old way is the right way. We tried modifying the matrix, and then using quaternions, and now we're back to modifying the model's matrix.

We also are finally paying attention to menus. We got in some artwork for today for JeffM to drop into the game. JeffM has been all but neglected since he's working on an unnoticed part of the game, the 2D stuff (HUD, Menus, etc.). But we all appreciate him taking on something he's not particularly fond of so that we can have a finished game.

6:08 PM 12/1/2004

A breakthrough today. JeffH has been able to use the RenderWare exporter, and finally got all the joints exported. In order to get them all exported, he had to specify the joint information in Maya and then export them. This means that our 3d artist doesn't need to split the model, but just that he has to export the subsequent models in the same way. He has animations and sub-animations working, using his models and his code (as opposed to the demo models and code). For now, he needs to export their animations. With the tech proving itself to us, the rest of the work is just implementation built around it. I've been trying not to give myself a heart attack over whether the whole sub anims thing would work. It's been a big risk putting time into something that we only _think_ should work. But as I've said, bad animations make for a bad game, and we can't have that. It was a risk worth taking.

Now that JeffH is exporting RW files properly, he can give Peter the corrected model for him to implement with Ragdoll Physics.

The HUD required JeffM be told what input he would receive in terms of what spells he would have, and how we would handle changes in health, and mana by stats.

I'm finishing up the basic scripted AI behaviors (runto, runaway, attack, circle) and am submitting that for integration tonight.

7:16 PM 12/2/2004

Today I went to class on an off day with JeffH to get some feedback on some animations. JeffH has subanimation code now, and all that remains is just getting the subanimations to work in the build. We thought it best to have JeffH write some theory code to test out the way he believed RenderWare would work, and now he needs to drop it into the build.  There is some issue that needs to be resolved with torso rotation whilst running though.

While at class I managed to sneak a peek at the Fikks, and at Rik Levins' new intro video. The former looks pretty creepy, even without the texture, and the latter looks most exquisite; I have never seen linked list code look so good.

Unfortunate news, the 15 minutes of game footage that took 1 hour to make is no good. The video only captured the upper left quadrant of the screen.

7:30 PM 12/3/2004

The artists are starting to respond to the Beta deadline that is this coming Wednesday, the 8th. I spoke with the Modeler, the Audio Guy, and the Level Editor almost consecutively today. All player animations are finished, as are all of our weapon models.

The audio is looking pretty tight as well … or sounding … as the case may be. I'm not sure if it was my overly detailed descriptions or Darryl Stephens' talent -- but for now let's just say both, hmm?

Erick put what we think should be the final tweak on the Forest Collision mesh, and gave us a base mesh for Nu'Ar Cave. In the interest of time, he's starting on the first floor, and if he can get to the second, he will. Otherwise, we'll make the exit back to the fighter's field be the entrance to what would have been the second floor. And no one would be the wiser … unless of course, they read my journal.

I start on gameplay code today. Party System is first, then Leveling, then the Ladder. Calls to the gameplay code will be made from the engine to determine damage, hit points, mana usage, etc. I already have prototyped versions of all of the algorithms in excel spreadsheets and Design Doc, but now I have to unroll the nested IFs and the pseudo-code into actual code that can integrate with the main build.

11:48 PM 12/5/2004

3 days remain before the Beta build is due. 15 minutes ago, we took some video of the game as it stands. JeffH finished the basic magic and ranged attacking. We all ran around in the fighter's field and killed each other with reckless abandon.

At the moment, the archers have a ludicrous advantage. A chronic backpedaler can kill just about everyone with a bow. Beyond that, melee fighting is actually quite exciting. When an attack is successful there is no confirmation of success. However, once audio gets integrated, we will cue up slice and pound sounds to allow the user to mark his progress.

Peter is scheduled to make a particle engine is as well. Currently, he is having issues getting RW's particle engine initialized. He still needs to actually write the particle transition functions, but once the engine itself is moving, the rest should fall into place. None of us suspect it to be terribly difficult. With that said, we have been wrong before, and are prepared for the case that it cannot be completed in time.

The gameplay algorithms I'm writing are coming along as well. I had a few problems with magic, as nearly every aspect was tied to only one stat. I now have proficiency determining speed and mana drain, with the Ch'i stat affecting the actual magic damage. This is actually more in tune with the rest of the system, where Strength affects Melee Damage, and Dexterity affects Ranged Damage. The stat and proficiency leveling systems were written months ago, and because of that, we've been able to modify the game and the systems to fit each other as we go along. This is opposed to just making things up at the end, as I am certain many others have done in the past.

JeffM has finished a loading screen, but we still have a number of menus which need to be written. I said I would help him finish the menus in the case that he can't finish them all in time, but I would still have to learn the format before beginning. Hopefully it won't come to that.

In all, I think we're all feeling pretty good about our game before beta. That's even without the special effects. Our core game and gameplay are all in before the actual deadline. Everything that I've heard from the instructors here has been a pat on the back and a silent "this game better rock" conspiratorial wink. I concur, but the enemy of success is complacence. We're in the homestretch, but I can't let the group rein itself in.

6:57 PM 12/7/2004

We put our Beta build on the test machine today and it ran with no errors. With the network down, we were forced to run the server alongside the client, which is not supposed to happen under any circumstances. I also got some finalized sounds and a game track from Darryl Stephens.

Also, I have the final Fikk Model from Rory, and it definitely looks creepy. He really took my idea in the right direction. I have to cut this entry short, I want to finish up the gameplay code so that we can get proficiency and stat leveling in by the end of the week.

11:06 AM 12/9/2004

Yesterday was Beta. I never think presentations go well. The oral presentation was even better than it was at Alpha, which I did like. The network (specifically the router) was being funny again. We couldn't connect anyone to the test machine because of it. 5 minutes earlier though, at our table, we were all fighting each other in the same game. One of the TAs told me that it goes from 5 to 0 in sixty seconds. We almost forgot to talk about the Bow and exited back to the menu, where the buttons stopped working. We displayed our combat system, but with no one to fight, it doesn't mean much. There was also some terrain rendering error that we'd never seen before. For some reason the right side of the path around the trees in the Endres Pathway looked like a crushed Dorito.





Diary: Through Final


Contents
  Introduction
  Diary: To Alpha
  Diary: To Beta
  Diary: Through Final

  Printable version
  Discuss this article