Hello from Team Abyss! ...Actually, it's just one member of the team here, but it's still going to be a we situation!

For this year's O2A2 Visual Novel Jam, we developed Goodbye, Emilia, a small tragic drama about a brother and a sister living in a world of despair. In a world where magic users are captured and killed for being a potential threat to the country's ruler, Emil (older twin boy) and Emilia (younger twin girl) find themselves trapped in a hopeless situation: either starve and die together, or have one of them distract the enemies (and, presumably, get help) so the other can live. Being the older twin, Emil tries to distract his sister by telling her a bedtime story and then leaving when she's asleep. And depending on how he tells the story, that plan may or may not succeed.

Now, let's get started with this postmortem, shall we?

What went right?

    1. Getting the story done early.

We started this jam a day after it began (mostly because we didn't really pay attention to the starting date, oops) and practically had nothing planned out. Compared to our previous game jam entries (Hellevator and Abyss), Goodbye, Emilia was a spur-of-the-moment thing. The story idea more or less came into fruition in the middle of lunch...and then it suddenly became this.

We finished the script on the first day of development in one sitting, which allowed us to focus on the other aspects of the game. One less problem to think about, right?

But it wasn't until the second last day of the jam that we rewrote the entire thing, and not until the last day that we re-implemented the narrative. Truly the crunchiest of crunch time writing!

    2. Having a clear direction with the GUI.

With the help of the very useful vector and stock photo website, illustAC, we managed to produce a rather decent-looking GUI, if we may say so ourselves. In fact, creating the GUI assets wasn't a big problem at all; we had a clear idea of what we wanted the screens to look like, and even added the Gallery screens from the get-go.

And then there came the other problems...dear Brahmi, the other problems.

How about what went not-so-right?

    1. Lack of drawing experience.

Because only one member of Abyss Team handled everything, they had to cover all the aspects of the game. And that meant producing the sprite, GUI, and music on top of coding the entire thing. And that also meant trying their hand at drawing a character without any particularly solid skills at drawing. And that also also meant figuring out how to draw variations of human expressions while being absolutely terrible at it. Not to mention the awful gradient colouring (we wanted to aim for cel-shading but had to prioritise time for other tasks).

Even up to the point where V1.2 of the game was released (post-jam), we were still trying to clean up some of the expressions to make it look...less weird...in the game. Too much time and effort was spent in drawing and colouring the sprite and its expressions. Work that a proper artist could've done in a few hours became a two to three-day endeavour.

    2. Overly ambitious GUI.

And so came the next daunting task: GUI replacement. Yes, we could've used the Renpy default GUI. No, we didn't because hey, customisation is cool! ...Famous last words.

XY-coordinates, screens, GUI bugs. Many hours spent figuring out why things didn't look right or work as intended. Although most of the GUI worked in V1.0, we spent three post-jam days just trying to clean up the GUI...and that led to more bugs, which we thankfully (maybe) managed to kill. Nonetheless, it was a lot of time spent just trying to make the GUI look as intended.

    3. Very last-minute audio production.

The Abyss Team member that handled the entire development is well-versed in audio production (music, sound design, and mixing), so they thought it'd be a quick and easy thing to do. They certainly didn't account for the last-minute decision to add full voice acting...which, incidentally, was recorded in one sitting. At least that was one thing about the audio assets that was quicker to get done with.

The audio assets were, rather literally, produced on the last day of the jam. This led to a lot of rushing around with cleaning up the VA lines and composing the music track (we ended up using a simpler version of the original track since it took too much time to do proper automations). And with all the lines voiced and the fact that they had to be implemented after cleaning them up, it was...a chaotic last day, to put it simply.

When did we finish composing the track? ...1 hour before deadline.
When did we finish cleaning up the VA lines? ...30 minutes before deadline.
When did we finish implementing the VA lines in the script? ...3 minutes before deadline.
You get the point.

Conclusions

There's a reason why game development is done in teams (unless you're one of those crazy geniuses who can handle all aspects of production by yourself and still can come up with a great product; you know who you are). Handling multiple tasks alone can be highly inefficient and time consuming if your skills aren't optimised to do certain things. The one Team Abyss member could easily complete the writing, narrative coding and audio tasks, but had a lot of trouble with the art aspects and GUI coding (mostly because it took a while to get the right look).

Nonetheless, it was a fun experience and a good exercise in managing time in-between real life responsibilities and work. The limitations of the O2A2 Jam was actually great since it helped us rethink the ways we can convey a powerful enough story in a short amount of time. If there are other jams like this in the future (where there are tight restrictions), we'd be very interested in joining them!

That's all for today. Thanks for reading, everyone!
--Abyss Team