A large aspect of the game contest was to test the scene system to make sure people could do what they wanted to do. 

This ended up working pretty good. Bugs were found, issues were made, bugs were fixed. Not always as fast as I would have liked.. but issues getting fixed is very important. 

We're in a really good position right now where someone can make a minimal project showing a bug, zip it up, post an issue.. and when we download it and unzip it, all we have to do is run the .sbproj file and it loads and shows the issue instantly. This feels really good from all perspectives, something we should lean into and make even better in the future.

There are still plenty of issues, and we're committed to solving them - but we're in a much better position now thanks to the contest (and the people who bothered reporting the issues)

Choosing the winnersattribute 'target' not allowed was hard. There were definitely games that I felt should have placed that didn't. Games I had tons of fun playing didn't place. I guess that's a sign of a successful game contest.

Ultimately I think that GUSattribute 'target' not allowed ended up picking the games that he felt could be released on Steam. Therefore the most "feasible games" won.

There is no doubt that the winners all very much deserve their trophies. The things they managed to do blew us away.

This was awesome - we're definitely doing this again soon.

The games are obviously suffering from a lack of players, so we need to find a way to get people online to play together more. We had some pretty good spikes, but there wasn't enough to have 24/7 people online. 

Actually, releasing s&Box is the solution to that, but that brings a whole bag of other problems. I think we'll probably try a middle ground option for the next contest, and then do a full release after that.

You know, playtesting these games is like Christmas morning. You see all the presents laid out, you pick one, unwrap it. Socks. So you pick another one, unwrap it. Socks again. So you pick another one, electric guitar what the fuck.

The stuff people managed to do was amazing. I know every inch of the scene system and I couldn't understand how some games had done what they'd managed to do. This system has existed for 2 months how the hell are people doing this shit.

In my most optimistic expectations, I thought we'd get 15 flappy bird-type games. That level of complexity, that level of depth. So the stuff people have managed to do has really impressed me.

I had a couple of moments of validation when playing these. You know, when you work on an engine like this.. all you see is bugs and issues. Everywhere you look you see things wrong, things missing, things that are sub-standard. 

So when you get to play with it for a bit, and nothing goes wrong.. the fog kind of clears, and you can see what you're doing, and that it works, and that you can make it even better. It feels good.

Early this year we took the decision to move s&box to a whole new way of making games. 

Previously, our game loop was source engine. The same as Half-Life, the same as Garry's Mod, the same as Team Fortress. Tried and tested. Although that has worked for 25 years, I felt that ultimately, over time I'd regret not switching to a simpler system.

As a result of this decision - we killed every game that was created previously.

Game Contests are something we talked about internally for a while and this seemed like a great time to start doing them. So we started building the underlying contest system and launched our first game contest.

We have things to do here, for sure. We need to bring party mode back and figure out a bunch of stuff about how to leave a party when you're in the game, party chat etc.

But holy shit - the new multiplayer system works. Someone creates a lobby and you join it and it just works. When they leave it just works. I know I shouldn't sound surprised but we played one game together for about 40 minutes before I realised it was multiplayer and was just working.

The contest started slow. The first entries were low effort or meme games. For a moment I thought it was going to be a disaster. 

So what was obvious was that people didn't really want to show their cards too early. We tried to encourage a bit more community interaction by introducing the news feed, where games could post updates. I think, ultimately, this was too hidden for people to care, but people started using it.

Over time as more games got revealed, I got a bit more relaxed.