Legacy:Jb

From Unreal Wiki, The Unreal Engine Documentation Site
Jump to navigation Jump to search

Chaotic Dreams Team Member

CUT Team Lead

Lead Coder/Mapper for ChaosUT

Projects for UT

ChaosUT

Popular mod that combined new weapons as well as started to branch out into other game types (KoTH). See: http://www.planetunreal.com/chaotic I was a new member of the team 1.5 years ago. Started as a mapper with some skinning skills. Then was promoted to team lead and lead coder. Yikes!!!

Plans for UT2k3

ChaosUT2:Evolution

The next gen of Chaos. Take the things that worked with CUT and improve or evolve them into something bigger, better more fun :)

ChaosUT Post Mortem

ChaosUT has had its share of success in the UT community. It did place in a few spots in the Make Something Unreal contest as well as being including on the GOYTE Unreal Tournament CD. But we also have had our share of problems as it was not such a smooth ride. Most of these comments comes from my experience being the team leader as well as a coder and mapper for the Chaotic Dreams team over the past 18 months. I just wanted to touch on some things that we could have done better and hope that other teams can learn and find ways to avoid some of issues that we suffered:

Key Talent Leaving

I am sure most teams that have been around for any amount of time will have to deal with this. With us it was a sever blow. Chaotic Dreams was a decent sized mod team coming off the Q2 versions (which was included on ID's Extremities CD). However the group wanted to work on both a UT version as well as a Q3 version. So the team sort of split into to areas with each "team" working on a Chaos version for that game. The UT version had a bit more support and we were able to get up to a GOTYE. However after that point some of the team decided that they needed a break. Our lead mapper just had a new addition to his family and wanted some much deserved R&R. Our only texture artist also was pursuing other interest in her Real Life and no longer could help out. Our founder/modeler/skinner just graduated collage and took a job working on the Y-project so he had almost no time to help. And finally the only coder also needed time off as he wanted to get his MS degree. So that left the once popular mod literally dead in the water. They all did not leave at once which was a good thing. In fact I was brought in to replace the mapper as I had been a beta tester for them. Like wise we were able to find a few more people to help all taken from our beta test group. As you can imagine we were still learning how to make a mod as well as our skills had not developed to a level that was close to the original team. This meant a lot of work for us new guys. I know that this is a pipe dream but having extra people or people that can overlap and do more than one thing is a god send. I was able to learn how to map and code. We also had another team member that could map as well as model. So we could get by with out a dedicated mapper. Not every team can afford to have people get experience in other areas but it not only makes the team stronger but it give that person a chance to work in another area which for some is a nice change of pace. Also networking can help to replace talent. Getting your name/mod known can help produce leads and draw talent in. You never know when you may need more help.

It also would have been a lot easier had these key users documented a bit more of what they did. Us new guys were not given any guidance or direction. The original coder did have a chance to give me all of the code files so we could continue with the code but a lot of content had to be recreated from scratch. While that's not the worse case it only managed use to take more time to get to a point were we could move the mod forward.

I know that things change and it will be hard for a team to stay together. But make sure you have a back up plan in case any key member needs to stop working due to what ever reasons. And if that person has to leave try to get them to document as much as they can as it will save a lot of time and energy. Also try to keep your core info some place where others can access it. You don't want one person to have all of the core data and if that person looses the data then the whole team may be in trouble.

Beta Testing

I like what Mychaeel said about beta testing and how that's a key component. At first we did not have a system to test our mod. So our beta testing consisted of a few of us playing the mod looking for any errors. While that's fine and dandy for some part it was not enough. It was my fault that we did not concentrate on it more. There were some bugs that should have been caught but were not. Also that method is not very efficient. We have since refined our method to test for certain areas and have broken up the work to each beta tester. This allows us to not only cover more areas but also spend less time doing so. It also helps to have good testers. A few we had in the past did not document very well the "when and where" when a bug was found. So it was hard to reproduce and fix (I was probably the worst beta tester we had). The current group of guys has really done a good job of providing all of the details on how to create the bug so I can attempt to fix it vrs spending hours to reproduce it. I would like to see our beta testers grow a bit more so we can continue to refine our methods. Beta testing is one of the most important parts of the mod making process; don't over look it!

Coding

Here is another thing that I should have know better. I thought some classes in assembly coding while I was in my graduate studies at school. I stressed to my students that's comments were essential. Yet I look back at the code I wrote and I hardly have any. What a hypocrite I am :( The original coder did not have time to comment his code and I was always working on it in a "fire fight mode" meaning I need to get the code working now. So I did not spend time looking over each and every line of his code. And with out comments I was lost in parts. I made assumptions and most of the time those were wrong. With the new code I did write I did comment it better. And no I don't mean every line needs to have comments. But looking at code with nothing commented is hard to do. It would have saved me a lot of time had we had some comments in our code. The new code we have written for the next version of Chaos is well committed and we will keep it that way.

Also at one brief time there was two of us working on code. There was one case we both worked ended up working on the same thing and had over lapping code. The problem is we lost some of my work when he did the code and vise versa. If your sharing code you need to have good communication. I would prefer to have one person keep control of one part of the code that they are responsible and the other keeps track of a different set of code that is their responsibility. Other wise sharing code is a nightmare...

Team Size

Chaos never really had a dedicated leader or any set positions. For the most part we were just a bunch of people hanging out having fun. However as the team has grown a bit it gets a lot hard to keep things moving. You have a difference of opinion on all sorts of things as well as some confusion. We have since tried to make some general listings so we can delegate the work to the right group. I agree this sounds like common sense but that's something we did not do at first.

Public news

During the time we had transition from the "old" to the "new" team we had almost no news or other public showing. I think that really hurt us. I have seen some mods draw in a huge following with out releasing anything. They do this by many different means. Now while I don't want to change ChaosUT to mod that wants to be popular, I do want to keep the fans more in tune to what's going one. Once every 7 weeks, updates to the sites will not get that goal done. We have since re-did the interface to submit news to make it a bit easier for us to keep the public up to date. We also want to start posting more concept art and development work. We have seen a lot of other mod teams post renders of their ideas in various stated of development. We are not making Chaos for us, we are making it for you and we need to work on that last part better.

I am sure some of the other teams out there have done all of these a lot better than what we did. And most of these are common sense. However I am sure we are not the only mod to suffer some of these setbacks and other issues. So I hope that we all can learn from our mistakes and make sure that we do not make the same mistakes next time. Thanks for your time, Jb


Mychaeel: Welcome to the Wiki. :-) Add yourself to the Project Contributors page, please.

ZxAnPhOrIaN: Hi Jb! :) :tup:

Tarquin: That's some great advice in your history of ChaosUT, Jb. I'm starting to wonder if your ideas from this page and Mychaeel's could be made into a general page on Mod Creation.

EntropicLqd: All good stuff there JB. For all that it might have been painful, you chaps succeeded where a lot of teams would have simply folded up. And Real Life is a right bugger as when the cards are called the "game" always loses. I'd be interested in some more detail on your testing strategies .. good testers are damn hard to find.

Jb: I'd be happy to help in any way if you think a page like that is a good idea Tarquin, Just let me know what I can do to help.

EntropicLqd: Have a read of Making Mods/Effective Testing and change, comment, refactor at will. It looks like Making Mods might become a reality. We shall see.

Vajuras: This was a very entertaining read. Choas UT has definetly accomplished a lot of great feats. Good to see things are back on track!

LoQtUS: That sounds like an acurate account to me ;) Allthough i really dont think i was a god send. heh