Are UDK and Unity suitable for a semester long group project by CS undergraduates?

Question

For the past two years I've taught a final year undergraduate course in game development with the main focus being a project in groups of 4-5. I've required that the students use the NeoAxis engine mostly because it what I found accessible, and it comes with a good range of demo games that the students can "evolve" to become their own games. Plus it works with Blender which helps with our budget. Overall it has worked pretty well.

This year I've opened up the choice, and in particular I mentioned Unity and the UDK as alternatives. I've tried both a little, but now that the students are starting the project I'm wondering it's wise to recommend these given I know less about them compared to NeoAxis. The UDK worries me more than Unity just because the UDK seems so fully featured that it's easy to get lost among the features, and the time frame for these projects is just one semester. (But, maybe I should be more worried about Unity.)

So, does it seem likely that undergraduates with relatively strong technical skills would figure out each of these engines (Unity, UDK) sufficiently during a 12 week semester to produce a "prototype" level game, similar to what's been done in previous years with NeoAxis?

Answer 1

I haven't worked with either NeoAxis, and only have a little experience with UDK, but I'm pretty familiar with Unity.

I just graduated last year myself, and as part of that I took a one month course in game development at DADIU - The National Academy of Digital Interactive Entertainment in Denmark. This was while I was taking my masters degree, but undergraduates can follow the course as well. Anyways, during this month we worked in teams of 10-12 (if I recall correctly) with 3-4 programmers and the rests spread out on artist, animators etc. We worked with the Source engine, with no prior training and all teams were able to create a fully working game within a month. Since then, DADIU have begun using Unity instead, and the results have been amazing! The production still lasts for only a month, and students still don't get any training, but people still manage to make some really amazing games, some of them even being featured at game festivals e.g. Puzzle Bloom and Flipside

So yes, I definately think undergraduates will be able to produce a "prototype" game within 12 weeks. However, it might be a good idea to lay down some restrictions on what they are allowed to do, so they don't loose track of the goal. DADIU in example, each year has certain rules which all teams are required to follow, e.g. the game can only be controlled using four keys and no mouse.

Answer 2

UDK does indeed have a harsh learning curve. It's all flowers and sunshine as long as you're using the editor, but there's no documentation on their changes to UnrealScript and all the tools and code samples around the web are now outdated. I've also experienced a good number of crashes with the old Unreal editor -- fewer with UDK, but I wouldn't call it stable.

I personally think it's a mess. I would have preferred them to keep UDK under wraps for another year or two until they did some QA on it, made sure it was all ready and released it with documentation.

I'm not a huge proponent of Unity but when it comes between the two, I find Unity to be much better, cleaner and more professional. But at the same time, I have very little experience with Unity. I am simply recommending against UDK from my prior experiences unless you want to stay within their editor. Oh, and don't forget things change every month; several months ago I had problems with compatibility in a game project when they made some pretty huge changes. That was time wasted.