Basic gist
Mizu is a 2D platformer where you use water on your advantage.

You play as a kitchen glass that is able to jump around, but also to take in water and use it to put out fires or to press down heavy swtiches in order to open up new paths. The only problem is that the more water you take in the slower you will move and the shortest you will jump, so you have to be careful and play smart according to the situation. To make matters worse, this game's world is full of nasty pitfalls and mad and violent forks that will try to stop you from reaching your goal.
You can play it right here: https://ldjam.com/events/ludum-dare/40/mizu
Development
Day 1
Since I live in Europe I didn't learn of the theme until Saturday morning, and to make matters worse I kind of overslept. Around 10:00AM I crept out of bed, checked the theme and got straight into the shower (is a lovely place to think about stuff). Surprisingly I almost immediately came up with the glass full of water concept. I also wanted to do a platformer focused on a simple mechanic and explore it in many scenarios, so I quickly started drawing and planning the scope and needs of the game.

After that I started coding up the basic platforming system and the map parsing classes, in order to be able to easliy create maps with the Tiled editor <3
I use HaxeFlixel and I'm quite used to it by now, but I ran intro some trouble when loading the tilemaps in tmx format, all because Tiled had a new version that doesn't use a simple reference to the image file you are using for your tileset. Anyway, figuring that out chew away some valuable time, but apart from that coding went by smoothly. I finished up around 23:00 and then decided to have dinner, watch a movie and go to bed.
Day 2
Woke up early and started drawing sprites, tiles and any other kind of graphic resource that I thought was going to be needed. Making sprites for the character was a bit of a pain, considering that I had to replicate any animation in every state the glass was, that is, for every amount of water the player was going to be able to carry I had to make a whole set of animations. Luckily for me I set on a really easy to animate character that only has two feet and a pair of eyes, so everything went faster than expected.
I also implemented a few things that I had left to do, like switches and moving blocks. I also decided that killing enemies wasn't any fun, so I removed that option and kept the game completely based on jumping, avoiding and your puzzle-solving skills, which I think was the right choice.
By 19:00 I started making the levels, which was a lot of fun and actually didn't take long, considering I had everything completely set up and I had a pretty good idea of what I wanted to do on each level. My initial plan of a tutorial level + 5 normal levels changed into 1 tutorial level and 3 normal levels, as I condensed some gameplay ideas into the levels I already had. Luckily no bugs appeared at any point and everything went super smooth, which is something I'm quite proud of (I have a long history of Ludum Dares when I'm fixing a game crashing bug at the last moment...).
By 23:00 the game was uploaded and fully playable online, which left me some time to start checking out the rest of games you people had been making as well.

Conclusion
Overall I'm quite pleased with my entry. I managed to deliver something that is complete, focused and is fully implemented. The mechanics of the game are clear and the game design does its best to showcase them. The game has a full graphic set with animations and also features SFX and music. Also, there weren't many bugs to talk about, and so far the game has received pretty good reviews. It's been a long time since my last attempt at a Ludum Dare (early 2016 I think) and seems like I'm back in my game jam shape :)
The good
- The game explores a single mechanic and tries to use it in as many situations as possible, while keeping the game simple and yet challenging.
- Graphics and music wise the game is complete and is not missing anything.
- I managed to do all this without being stressed or having to cut out on sleep time, overall, I did it all in a day and a half's work.
- Amazingly, people seem to be enjoying my crappy music!
The bad
- The tileset should have been bigger, allowing for more diverse and rich enviroments.
- There should've been indications on how much water do you need to push each button.
- I should have made more levels, but had to condense the ideas I had in fewer levels to keep the game focused.
- Had to deal with some nasty Tiled vs HaxeFlixel crap.
- Windows and Linux versions are missing, but HaxeFlixel seems to be reluctant about the matter since I upgraded to OS X High Sierra.
Anyway, that's all for today.
If you want you can check and rate Mizu right here: https://ldjam.com/events/ludum-dare/40/mizu
Thanks for reading!