On conception, Flickerlight was a much larger game. Here's a short list of things that didn't make it in: - combat and swords. - teleportation network. - three times as much dialogue as we ended up actually getting done. - an entire audio track that we didn't have time to write the music triggers for. - these cuties (thanks to Valintone for pulling off the extra work, and sorry they didn't get used!):

This one is clearly scheming something:

LOOK AT THIS ONE HE'S PERFECT I'M SO SORRY YOU DIED OF SCOPE CREEP YOU DESERVED BETTER

We'd never done a game with dialogue before, and I think that was our main - well, I hesitate to say 'downfall' as I'm pretty happy with the end result given the time constraints. But we definitely had less experience in scoping this sort of thing than we do our usual offerings, which tend to be 2D physics sims with scaling-difficulty-based game loops and some emergent behaviour. This was a whole different category of beast.

The biggest difference was in...let's call it prototyping. With a physics sim, you usually have a playable game loop by the end of day 1 and the rest is balancing and polish. Here, the core game loop didn't really become playable until well into day 3, at which point all the art and dialogue and resource management logic finally started clicking into a cohesive whole.

Having dedicated artists on the team made a big difference - but again (sensing a theme here) it was something we weren't used to and didn't really know how to make the most of. As much as Flickerlight isn't really what you'd expect of a 'game jam game', I'd love to try this sort of thing again.

Speaking personally, there were three problems with doing dialogue without prior experience. The first was learning the codebase and dialogue system (as I am not the main mechanics programmer on the team), and the second was estimating the time the work would take. The third problem is less directly relevant - as I am the team's jack of all trades, I often failed to account for how much of my allotted 'dialogue time' I would actually spend on other things, such as sound effects and minor mechanics tweaks.

Those who stumbled across the dialogue path pictured above may have noticed that there was a LOT of backstory, compared to the remainder of the dialogue. Judging optimal dialogue tree depth and levels of recursion was very much something I had to learn on the job, especially once the task management system was finished and every character with a 'complete' dialogue tree suddenly had to have job system compatibility edited in.

The one good decision I made this time around was to do the music first, rather than at 6AM on the last day (Australian time - AEST). This took the pressure off - but it also gave us a moody audio track to use as basis for the theme and mood of the whole game. Personally, immune as I am to listening to the same thing for hours on repeat, I played the main track for the game in the background throughout most of the rest of the jam.

Other than this, there's not much else to say. I worked with some new people, who were a lot of fun to build and cooperate with, and some established teammates, who pulled their weight admirably. I wouldn't have it any other way.

Looking forward to the next one :)