When Orcs Attack artist post-mortem

Right, there's already two postmortems for "When Orcs Attack" (Jons and Johns) but I figured I'd do one from the artists perspective as well.
(well, okay, Jon bitched at me to do it so here we are...)

For all of the back story etc make sure you read Jons post. The gist of it though; Make a game in 3 months. Or more precisely, make a game in 12 months but with a "deadline" that should result in a separate game of it's own every 3 months.

Instead of doing the regular "what went right" and "what went wrong" I'm going to go about writing this slightly different since most of our decisions had both their advantages and their share of problems. Lets just get on with it shall we?


Let's do the time warp....
Time is obviously going to be a pretty big issue when you're doing a game in 3 months no way to get around that. This will be from my perspective though so I'll just go with that; As Jon mentioned in his postmortem he didn't get me started on the project until a bit too late. With the next game I'll be working on it from the start. Let's just leave it at this; Some things were cut, some things were rushed, a lot of coffee was consumed.


Premade content.
At first Jon was pretty set on using as much "premade content" as possible. Personally I saw the dangers of this right away. It DID (and do) work out great for quickly prototyping and getting the game to a playable level very fast but a lot of the content wasn't optimized properly for our needs, and not all the content we needed was available to license. So the bottom line was that we would have had to optimize a lot of models as well as producing more art that would artistically fit with the content packs.

From my experience it's quite often simply faster to just create it all from scratch in these cases. If not faster so at least roughly the same amount of time and you will end up with an art style that is yours rather than that of whoever made the content packs you started with.

Jon was a bit hesitant towards the suggestion of throwing out all the premade content at first, but quickly changed his mind when I started producing the assets.

Bottom line; use content packs for prototyping, don't do it for the final game.


Plans and anti-plans.
Planning is certainly one of those things that are both a blessing and a curse. In all honestly our planning really wasn't particularly extensive and we were mostly "winging it", both on the programming and art side of things. Most likely the only reason this even worked at all was because I've been working with Jon for quite a while on "Harmonic Convergence" and he trusted that I would figure things out.

Due to the lack of planning I was given a lot of freedom with the art (well "a lot" may be the wrong word. More like "absolute" freedom), as I said; both a blessing and a curse. The good thing with no real directions as to what art is needed is you can choose to do things that you know will be quick and efficient to produce, the drawback is of course there's some time lost to figuring out what to do.

It worked this time, but hopefully we'll have at least a bit more sense and plan slightly more in the future. Some flexibility in the planing is needed when you do a game in 3 months due to features being changed or cut, but having a rough idea of what's needed from the get go makes it easier to schedule your time.

Extremely quick, dirty and butt ugly sketch of the walker robots in "When Orcs Attack"


Tools of the trade.
The choice of using Unity as our 3d engine certainly contributed to the game getting done in 3 months, that and the nice fact that when you have your game in Unity is a no-brainer to export both mac, windows and web browser builds. There's certainly been a few issues and annoyances along the way but nothing more than what I'm used to (I've been working with all sorts of strange engines over the years, there's always SOMETHING that require you to jump through hoops to get it all working properly).

One of the art related things with unity that I ran into was the use of FBX, basically; the 3d studio max FBX exporter isn't that particularly clever. I animated one of the towers in the game using a variation of IK and noise controllers (this particular animation was later cut from the game for completely unrelated reasons), the FBX exporter really didn't like this and instead of telling me that it simply ignored the animation data. I ended up having to write a maxscript that would sample the animation data at each frame and convert the IK and noise animation to regular transformations, shrug.

Another, very big, issue with unity is the fact that it really doesn't play along nicely with SVN (or any available version control system for that matter). We had been using SVN exclusively when working on "Harmonic Convergence" and this made our workflow extremely smooth, I could add an art asset to my copy of the game (as well as write any script associated with it) and it would instantly be in the main repository, or update any art asset and it would instantly be available to anyone working on the project. Without this things just take way longer, works way less efficient, there's all sorts of weird copies of the project files around and a lot of work that has to be recreated several times by different people. Definitely a major annoyance. Hopefully we'll be able to come up with some smoother solution to this in the future.

In conclusion.
There's a pretty simple conclusion here, making a game in 3 months is more than doable. It does require the right people, the right tools, a lot of coffee, some money and a certain degree of insanity but hey; It's possible!



IGC
Anyway, I see you all in Eugene next week. I'll be arriving Saturday evening (but I'll most likely be pretty dead by then after 20+ hours of flying and simply crash), so if anyone is around Sunday morning let me know and we'll go get some breakfast. =)

Also of course if you have anything particular you want to discuss with me at IGC, just let me know. (either there on the spot, or send me an email before hand).