I think I’ve been talking about Edu2 a lot already and I don’t really feel like giving a huge overview again - probably I will end up doing so nonetheless. Edu2 is finished, over, done. I’m not going to touch it again - ever! That said, I think there are a few nice things and lots of bad things to say and tell about it. I usually paint bleak pictures and I won’t stop here, because I try to be honest with the negative side of things.

I guess this is going to be a bigger postmortem because there are many different topics that I should mention to give the reader a fair impression.

To begin with, a few links first for those interested in Edu1 and Edu2:

Interesting fact: google for Edu2 and you’ll find a project that is almost about the same things that we worked on - except it’s 2D only.

First let me tell you what Edu2 is about: Imagine Second Life, Facebook, some uni/educational content and mini-games (flash stuff) merged to one product. That is Edu2 from what I could figure out of the business plan and my work on the prototype.First observation: this sounds like a huge project - too huge for one university term. Thus we only worked on a ‘prototype’. Second observation: it sounds like a project that is a tad bit megalomaniac and doomed to fail because the market is kind of saturated, too. I, honestly, think that this will be its likely fate, if anyone should ever try to develop it.

The project was actually about creating a business plan (half of the team) and creating a prototype (the other half), so I don’t think it’s a bad idea, because social networks really are big nowadays and experience doing social network stuff (even if it’s only a very raw prototype) can come in handy maybe. It’s mostly useless if you think that it’s helping you get a job in the game industry because it’s a very different business area and at most the prototype team may have learnt something valuable because it was working with mostly standard game tools.

One really funny note here: We visited Ubisoft Germany and EA Germany and presented our project there. At the time being there, I was somewhat excited to get an opportunity to talk to some important people there. However, I wonder: why did we even go to visit them? What do they have to do with our project? Why would they be interested? Obviously, they are not - except maybe EA, which we all know, plans to have a go for total world domination with unoriginal, corporate games (this is supposed to change though, so thumbs up) - so I question the use of the whole visits.

Had we got a simple but interesting game, it might have been of more use and we could have possibly pulled of what I would call a “Portal” and actually present something convincing that would have been a true first playable prototype of a game. So the first big point that deserves mentioning is:

A good project needs a concise and down-to-earth vision behind it.

Edu2 certainly didn’t have one, because no one really had an idea of what to do or rather everything was pretty blurry. We had a few docs with very rough but mostly too vague ideas and the problem was that decisions could hardly be made because, of course, you had the big real thing looming behind the prototype. In the end everybody or rather tiny groups of people who talked about it decided from case to case without lots of deeper thoughts behind it simply to get things done.

Communication in general was pretty bad

Really there was little communication between the different groups and I, at least, had very little contact to everyone else, so it was very hard to involve the team in decisions - since we were the coders, we pretty much had control over the core functionality, because our level designer didn’t have that much freedom: their task was to extend and clean up the original Edu1 campus level (a level consisting of a virtual university campus) and add gecko terminals now and then. We had a few fun ideas the night before the presentation, for example:

Adding zones that would open a gecko window with specific content - e.g. Amazon in the store or the mini-game selection window when you are near a gaming console..

Those were ideas that actually were really interesting and could have added a lot to Edu2’s ‘gameplay’ but sadly time was a very constraining factor and in the end they didn’t make it in, even though the code was pretty much in place. Additional trigger brushes would have been needed, but there simply wasn’t enough time to recompile the map. Clearly, we could have come up with those things earlier if we had talked and discussed things more.

The only person I was talking a lot to was Frank ‘res(2k)’ Richter, a pretty damn good coder, who is currently working on the free and awesome Blender/CrystalSpace ‘Apricot’ game in Amsterdam. He did an amazing job with offscreengecko, a gecko client that makes it really easy to include gecko (XUL) in games, and the Edu2/DarkPlaces code we worked on in general - DarkPlaces is the name of the engine for newbies to this blog.

We talked a lot about things and it was great to have someone on the project, who is at least as experienced as I was and quite certainly a lot more experienced than I am to be honest. A few things went wrong there nonetheless:

  1. I was chosen to be the coding lead between the two of us. Since there wasn’t anyone else, I don’t think, I really had much to do as “lead” and that there was a huge difference in what we did, but I tried to take responsibility and keep bugging him to work on stuff, question him about things that we needed to work on, try to decide on the tasks we should do next, etc.. It was a team process and I hope that he isn’t too angry with me about it :-/
  2. Again and again and this time, too, I’ve been way too protective of my code and way too restrictive and critical of other’s code. When someone changes things in my code - that is code I’ve been working on - and I consider the changes to mess up the code (design) or be problematic - that can drive me nuts pretty quickly. So I guess I’m quite bad tempered when it comes to coding. I’m pretty glad that I only frustrated Frank once for real and I’m still very sorry about it and I hope that I’ve learnt my lesson now.
  3. Even though I’ve tried my best, work was sometimes distributed awkwardly resulting in conflicts and things being done twice - although those few times I don’t think that time was really totally wasted but it certainly wasn’t used as efficiently as possible, which is bad because we both had other, more important things to do.

Another fun fact: no one has told our self-taught mappers and modelers how to do so effectively. Edu2’s only map apparently had compile times of 20+ hours at the end of the project and we had lots of bloated models with ridiculous triangle and vertex counts without any real justification for it.

I’m not sure I should mention this, but we had a lot of trouble with our only character modeler Jarno Müller, who, to put it in a nut-shell, wasn’t a least bit as professional as he claimed to be - considering the fact that he seems to own his own company that does gfx contract work. I’ve already ranted about him in a past blog entry, but I think he deserves to be named, because otherwise the way he behaved and partially crippled the project - we had to use standard free Quake3 models in the end for god’s sake - wouldn’t make a difference and the incredibly unprofessional way he acted and left the project (by simply not replying to emails at all in general anymore and not participating seriously in any way before either) pissed me (and others) off big time. He was offered help multiple times in the beginning to try and see how to export models from Cinema4D into a format Darkplaces supports but didn’t answer any emails and then a week or two after we extended his deadline to produce one model at least he sent us a model we couldn’t open because he used the latest C4D version and there was no converter to be found quickly - and he, of course, demanded that we tell him how to export the model to DP and he didn’t really adhere to one requirement that he was told from the very beginning (Quake3 multi-part model, easy support for different skins, multiple low-poly models - WoW-style).

Anyway, to sum everything up, here’s an overview of some of the pros and cons:

  • Pros

    • I have meet some interesting people and also made a few new friends (hopefully).
    • I finally worked again on DarkPlaces, which was nice, because I have only been working with GarageGames engines for the past two years.
    • We pretty much started from scratch game-code wise and had Edu2 actually been a real game and had there been a FPP in the end, it would have been an ueber-useful experience. But even without that, it was pretty interesting.

  • Cons:

    • I really have lost a lot of time with the project that I really could and should have used for other things.
    • The final prototype really is mostly useless, except if I get myself motivated enough to port back some of its features to Nexuiz..
    • The prototype hasn’t been actually officially used for anything else than a small video.
    • Thus I can conclude that it was a terrible waste of time

  • Worth mentioning:

    • The crunch time in the last two days was really fun and sleeping ten hours in three days says it all.. It was probably the most fun I’ve had in a long time (before visiting SplashDamage again).

I think that’s it for me.

Cheers, Andreas