Burning Down Assets

January 17, 2020

There are many methods and tools available for managing software projects. These have nothing to do with creating the software, instead they’re used for planning and scheduling purposes. One such tool is the burn down chart, which basically lists all the development tasks left to do over a period of time and is used to gauge how development is going (there’s a little more to it than that, but that’s the gist of it). Prior to the start of the new year my main focus on Sleuthhounds: Cruise had been on scripting different scenes and scenarios into the game. This is a very fluid process, equivalent to writing a prose story. Given that stories need to be as long as they need to be to be told, it’s not really possible to use a burn down chart to determine how much work remains. By the very nature of being a creative, additive process it kind of defies the use of a burn down chart. BUT...!

Now that all the scenario scripting is done, now that the story is complete, things are very much different. The boundaries of the game are known now. It’s possible to look at it and ask questions like how many backgrounds need to be illustrated? How many sprite animations are needed? How many dialog lines? How many sound effects? Development has now shifted into a much more quantifiable phase.

The consequence of all this is that it becomes easier to see how much work remains (a lot!) and to better determine if that work is being done to schedule or not. And for that, a burn down is very helpful.

At this stage, the highest level of breaking down the remaining work to do on the game is at the room level. Each location players can visit in the game is a room, from a technical standpoint. It doesn’t matter if it’s showing a wide shot of the exterior of the cruise ship, a cavernous theatre, a claustrophobic cabin, or a close-up of a shipboard newsletter. From a technical, implementation point of view, each of these is a room.

Cruise comes in at a hefty 185 total rooms. Of those, 79 are walkable rooms and 106 are non-walkable rooms. A walkable room is one in which the game characters can walk around as opposed to non-walkable rooms which are typically close-ups of documents, scenes for Homes to deduce, or standing characters for Ampson to interview. Walkable versus non-walkable is an important distinction because walkable rooms require additional assets that non-walkable rooms don’t. For example, a walkable room needs a data file to define the areas of the room that the characters can actually walk on (you usually don’t want characters being able to walk up walls). Walkable rooms also require depth information to indicate that if a character is standing in front of a table in the room then they actually show in front of that table. Or they show behind the table when they move there. There are a number of additional assets like that that walkable rooms require.

When working on my games I keep track of the time I spend doing different tasks. This is invaluable for estimation purposes. For example, from previous Sleuthhounds games for all the background related assets for a room (the background image itself, the depth data if any, the walkable area data if any, etc.) it takes about four hours to complete. As well, once the primary scripting is done, which is the state at which Cruise is now at, on average it takes another four hours to pull all the pieces together and completely finish a room.

So, with 185 rooms and an average of 8 hours per room it becomes possible to plot out roughly when all the rooms will be done. Working five days a week on the game and throwing a few hours in on the weekend I’m targeting 6 rooms per week, which works out to 30.8 weeks (call it 31 for convenience). That puts me into August of this year for the game to be mostly done. There are a few odds and ends outside of the rooms that will have to be dealt with and testing and so on. However, for now I’m just focused on the rooms.

To help progress things (and just for the sanity of seeing actual finished assets in the game) my current strategy is to finish a simple room in the morning and then spend the rest of the day on a more complex room. The current complex room I’m working on is the main atrium, which is comprised of four different floors and so production wise is actually almost equivalent to four rooms instead of one (I have other rooms in the game, like the passenger cabins, that reuse the same background and much of the same scripting so I’m considering that a wash at this point). For the simpler rooms, I’ve been working on the various character interviews Ampson can conduct. So far, I’m right on track for finishing rooms off. Hopefully, I can stick to that average of six rooms a week over (or better) over the longer term. Wish me luck!