Howdy! Adhesion from team RADMARS here with our postmortem analysis of our fancy LD27 effort SPACE JANITOR: CUSTODIAL MARINE. But first, goodies!
And for postmortems, first up we got spacemars:
Ho hO! I’m really happy with how the sprites turned out for this game. The janitor especially. I feel like I spent too much time on them though :(. The envrionments (except the mainframe room) all game out a bit lower standard than i wanted them to be. But overall really pleased with the whole entry.
The music is stellar as usual, and I think we did a good job shortening up this game to be just about the right length. Our previous entry Tessitron was a bit too short, I think Death Death evolution was too long, Its really hard to find that sweet spot where your game mechanics are interesting, but don’t feel spread too thin.
I agree with emarcotte about moving away from mellon for our future projects though, Its an awesome tool, but can be very limiting (and stressfull on our programmers *_*)
can’t wait for the next LD
And now emarcotte’s:
I was a little sad to go back to MelonJS, it gives us a lot of tools, but it also is so… confining. I think Tessitron was our best effort by far as it was the most free-form. Regardless, Melon gets the job done even if we end up having to keep a stare at its code to figure it out.
How many times will we fix the same bugs? Hooooly !@#$ — there needs to be a libradmars-melon. Between Brink, Escape, DDE, and now Janitor we have like probably 25% common code, maybe even higher if we actually took a solid look at how it all works. Let’s rewrite the level fade again! We only have 20 minutes left, why not?
One of the things we need to try and do more of early on is the rapid feedback in testing. My most productive times for me were when I could say “Test plz???” I’m guessing this doesn’t apply as much to the creative parts of the project (music, pixels, etc). Perhaps I am just crappy at working in isolation.
I am glad we got something nifty out the door given how little time I had to spend on it. I will try to not end up coding from a friends house again. It makes things complicated. Everyone did great in their role and while the final product feels a little less polished than previous efforts it definitely has more to it in a lot of ways.
And lastly adhesion’s:
Postmortem! This jam went pretty well but there were certainly some rough edges. MelonJS is a very comfortable platform for us now, but we usually end up going pretty conventional which is frustrating. I always want to make bizarre abstract games but it’s so hard to come up with good ideas like that when the theme comes out. I think the thing we have to remind ourselves is, melon isn’t really limiting in and of itself but it’s so easy to do a platformer (or top-down game like BRINK) that we end up falling into the path of least resistance. Next time we have to stretch our conceptual brains a bit more or maybe allocate more time for brainstorming on the first night.
Development went well of course, though there were a bunch of annoying bugs, especially towards the end which took away from playtesting time. It’s frustrating – even though we’ve used melonJS so much we always end up redoing work we’ve done before or running into the same bugs repeatedly (level change, collision, animation…). I’m really itching to collect some common code to make this easier next time (maybe even do a proper post-compo version of my first solo LD game with it too!)
Making the music was great as per usual. I managed my time pretty well (did most of the music on Saturday night) so I had the rest of the time to do playtesting and bugfixing. This time for the music I used the same arrangement (same synths, drums, etc) for all the songs and did everything in the same project file, which made everything a bit more cohesive (though with a little less variety) and saved me a bunch of time. I managed a neat trick with iZotope Iris to make SNES-style sampled instruments which was awesome! Definitely gonna use that again. Plus, doing the sound design was great – every time I get deeper into it and every time it just gets more rewarding and more fun. LASERS!
Overall, yet another successful Ludum Dare from team RADMARS! Expect something wonderfully bizarre from us next time! (I hope…)