I want to briefly discuss core scenarios this week. These thoughts are inspired by a chat I had a few weeks ago on level design with a friend. I’m a relative beginner when it comes to level design. I can break down a level and analyze what works and what doesn’t (and why). I understand how good levels work, but I have trouble building my own.
The idea of core scenarios is to create unique and isolated elements of fun gameplay. An example of this would be using your super jump to get over a wall you would not be able to clear using your normal jump ability. Or picking up an object and throwing it a target.
So, one of the processes for creating a good levels is to try out many different scenario types. You then filter out the bad ones, and iterate on the good ones and create a few more. Eventually, you will end up with a key set of unique fun scenarios. Once you have that, it’s easy to mix and match different scenarios to create a wide variety of levels. Most games are built on that concept. Mario, for example combines running, jumping, hitting blocks, avoiding enemies and stomping enemies as core ideas, and they build incredibly rich and fun levels around these concepts.
So that’s what we’re currently focusing on. Trying out different scenario ideas and keeping the ones that work.
On another note, I found out that we didn’t get into PAX 10. I didn’t receive an email from them, but I managed to work it out when the list was announced and our game wasn’t in there. I’m expecting to find out about IndieCade soon.
Work Done Last Week
Some more gameplay improvements, particular in the shooting and jumping abilities.
Undo/Redo functionality for creation and deleting of objects: On the first “undo support” pass a few weeks back, I only added support for transforms and scale, as I knew the creation/deletion was a bit trickier. It took a bit of time, but I came out with a better codebase then when I started, fixing a bunch of bugs in the process, and cleaning up a lot of code. The major hold up was how to integrate it with the component system. That’s done now, and I’m happy with the result. Adding support for new components will continue to be very easy.
Improvements to the editor, and how we save the state of different objects: We have some very dynamic elements, and it is a pain to make changes to the level without upsetting the setup of these dynamic elements. So I have split apart how the different aspects of the scene are saved. It is now easy to store and remember the state of active objects, and to edit passive scenery without destroying the rest of your work.
More work on the new site, mainly familiarizing myself with Joomla. Nothing shockingly new to report.
Work Planned This Week
More gameplay experiments, more scenarios, and more testing. Pretty much iterating on these three points all week.
Oh and I’ve been trying to track down a really annoying memory leak in the physics simulation. I’m already 3 days (well, evenings) into trying to track this bug down. It looks like an epic bug already. And it’s in the most complex part of my code, the core of the simulation. Fun times ahead.
