Of crabs and processes
Scared of our power to procrastinate, we felt the need to structure our work in a way that pushes us forward. After all, the only technology that you need is deadlines, and we believe that when it comes to making videogames that's no exception; actually, when it comes to making software, videogames are probably one of the most dangerous places in this regard (but making videogames is not just software!).
For humble software engineers like we are, "making videogames" is an enormous space of ambiguity, choices, and unknowns that doesn't fail to send me into "crabbing" mode at least once a week. I'm sure somebody has already abused the term "crabbing" for this purpose, but with it I mean the continuous falling from one rabbit hole to another, lateral movements into features, choices, or tasks that you are taking too early, and so on.
To keep going with the marine theme, we'd love to be dolphins and blaze forward, but for now, we'll be happy to be turtles busy with going straight. We're sure that what we came up with will be a surprise to exactly nobody, it's just a soup of some process management frameworks after cutting off the fat to stay lean. After all, it's just the two of us and we've got the luxury to skip some performance rituals (daily scrums cough cough).
We aim to have three things: 1) deadlines, 2) a short to medium-term predictability of where things will be going 3) a way to review our work to adjust our trajectory. With this said, let's lay out the framework.
Cycles: we'll (unsurprisingly) keep going with 2-week cycles. At the end of every cycle, we'll have a review of how it went and some planning adjustments, keeping an eye on how things are going with respect to the current phase of development. Nothing surprising here.
Phases: we've divided the development process into 4 phases: ideation (coming up with the game idea plus a prototype), pre-production (developing the core systems of the game, a vertical slice of the whole thing including CICD and what not), production (completing all features, followed up by filling in all content) and, finally, a release phase (wrapping up, QAing, releasing across all distribution channel).
Each phase will have the following
- a required reading list, i.e. preparatory materials that we gathered in phase
$CURRENT_PHASE - 1 - a preparatory doc that lays out the planning of the phase, to be specified at the beginning of the phase and adjusted as we go, to have a longer view than 2-week cycles
- an expected timeline (we are going with 3 weeks, 4 months, 11 months, and 1 month respectively; and yes, that's very optimistic)
- an expected state of things at the end of said phase:
- the game itself
- art
- marketing, community & web presence
- user awareness
- metrics (how many users have played the game so far? how much feedback have we got? Etc.)
- post facto, a review document of the phase, because cycle reviews aren't enough
That's pretty much it when it comes to the high level of things. No plan survives first contact with reality, so make sure to subscribe to be informed about how we are screwing things up 😇!
J&K
P.S.: bonus picture from Ukraine!
