Witchfire Autumn 2020 Update – A Tale of Two Milestones

Two important milestones behind us, one of which is a story of happiness, and the other a warning and a lesson.

Going with the idea of altering between making systems and let’s have a new playable, after a few weeks of the former we’ve decided to make another internal demo.

To be clear, it’s never a real demo, it’s always just something that looks and plays decently enough to be shown, in theory, to an external person. It’s just easier for us to call it a “demo” rather than “a theoretically presentable build focused on select set of features”.

We gave ourselves two weeks for that. The idea was to nail our core combat loop, and to see all the hard work, especially on the AI and animation logic fronts, finally come together. Stuff like this, where hit reactions are a mix of AI, animation and physics:

Wait, wait, wait… “Nail the core combat loop”? Aren’t we showing bits of gameplay for years now? Was this all fake or something?

Nah. We never fake anything. But it does not mean we were 100% happy with the lowest level core combat loop, the famous “30 seconds of fun”. What many people do not realize is that making a good shooter is incredibly hard. It all looks deceptively easy: give the player a gun, put some bots on the scene, bam, slam, thank you ma’am. But in reality, turning all that into a highly satisfying experience is a mountain of work and thousands of little things that all need to synergize.

So what we had before was, to us, okay, but not great. And we cannot afford to be okay, we need to be better. There were moments of “this is it!” – I mentioned them in some previous posts – but we needed something highly consistent.

We put together a plan, worked hard for two weeks, and…

*clouds split, a beam of light appears*

…it happened.

All these months, no, years of work finally paid off. I was spawning a wave, playing the 30 second of fun over and over again for hours. With a few different enemy types, and in various set-ups. Props to the team.

I think the best story on this I can offer is that right after the milestone, one of us, a man who is our irreplaceable Sanity Slap in the Face, The Harsh Judge, The Savage Reality Checker, called me and said:

Hey, so what the fuck happened exactly in those two weeks that the game is suddenly good?

Riding high on this wave we’ve decided that the next milestone will be focused on polishing the fragment we chose for the build. Let’s make it all prettier, let’s get to pre-final quality!

The milestone looked very much doable. After all, we already have a lot of stuff in a good place, so it looked as only a matter of putting final touches to a particle here and animation there.

Not only we have not managed to do it all, but we didn’t finish the milestone the next week, and even today there’s still a bit left to be done. To be fair, some thing were done on time, nice and dandy, but some took way longer than anticipated.

That should not have happened. We know what we’re doing …right?

The problem is, despite everyone working from home for months now (thanks, Covid-19) we have not anticipated how severely it affects those periods of development that highly rely on constant communication.

“Final quality” is one such phase. It usually takes place when you’re making a big event demo, like E3 demo or Game Awards trailer, or at the end of the project. During such periods a studio, does not matter if it’s five people or five hundred, resembles a hive except all bees are on crack. It’s all instant communication, with hundreds of questions asked and split-second decisions made.

But even with great communication tools like Discord, things just aren’t the same when you’re not in the same room. What would take an hour of back and forth can turn into two days, especially if due to the work-from-home your team members slowly drifted apart into different time zones. I mean, this can be highly advantageous, like when during the night one of us does assets that then another grabs them and works with them in the morning, but it is definitely not optimal when going “Final quality”.

The lessons for us were:

• We need to be ready for the fact that the true “Final quality” phase will take us longer than we thought it would.

• In order to remedy the situation somehow, we need to introduce core hours for the entire team.
  
  “Core hours” is basically a few hours during the day when everybody’s working and is available. An example could be core hours of 14.00-18.00. Someone can work 10.00-18.00, another can do 14.00-22.00, but they’re both here and present 14.00-18.00.

Unless, of course, the pandemic will be stopped next Spring or Summer and we will all get back to the studio. That works, too.

Right now, we’re working on a new, pretty challenging but super exciting milestone. We came up with the design for the core flow of the game (so it’s not about “30 seconds of fun”, but rather “30 minutes of fun”) and we’re prototyping that. I say it’s challenging because while not ground-breaking, it’s not something I’ve seen elsewhere. So we’re in terra incognita here, and that’s exciting but a bit scary too. Luckily for us, we do have a safe, true-and-tested back up plan, so definitely the exciting part is more present than the scary part.

Question of the Week

This is me patiently waiting for more Witchfire updates. I'm still super excited for this game, ever since the reveal! Keep up the good work @TheAstroCrew! I hope development on the game continues to go well. pic.twitter.com/L6qdphK3sm

— Nathan (@Newtzee18) September 17, 2020

It does. Actually, the fact these development diary entries are less frequent means we’re busy. That’s a good thing!

I do want to write more often but I’m truly swamped. Things were much easier when we had less game but now it’s in full production and I work on it days and nights – there’s still a lot of things to design and solve and playtest and organize.

I’m not complaining. Like this guy, I live for this shit.