{"author_name":"dylanwolf","cat":"LD #35","comments":[],"epoch":1461023520,"likes":1,"metadata":{"p_key":"6376","p_author":"dylanwolf","p_authorkey":"20078","p_urlkey":"41810","p_title":"LD35 Post-mortem","p_cat":"LD #35","p_event":"LD35","p_time":"1461023520","p_likes":"1","p_comments":"0","p_status":"UPD5","us_key":"20078","us_name":"dylanwolf","us_username":"dylanwolf","event_start":"1460678400","event_key":"33","event_name":"LD35"},"text":"

Here’s a quick rundown of the ups and downs of my compo entry, Shifty Shapes<\/a>, which was written in Unity.<\/p>\n

What went well<\/h2>\n

Concept<\/h3>\n

Usually I make some notes about the top themes in each round of voting, but this time I went in cold. (Honestly: I have a few different game ideas floating around in my head right now, so I was open to doing something off-the-cuff.) I had all of the rules for the game written down within minutes of the theme announcement.<\/p>\n

The sliding concept was inspired by a board game called Ricochet Robots<\/a> that I played in analog gaming at Geek Media Expo, plus standard match-3 mechanics which I’d already figured out an algorithm for in Ludum Dare 30<\/a>.<\/p>\n

Animation<\/h3>\n

I knew I wanted this game to have some nice visual effects, since I envisioned it as one of those shiny mobile puzzle games.<\/p>\n

Fortunately, there weren’t a lot of moving pieces in the concept, and I started building the effects early (even before I’d replaced the placeholder art).<\/p>\n

Animation is something I don’t tend to think about (or I think about so late in the jam that it’s complicated to implement), but a little bit seems to go a long way.<\/p>\n

Bouncing UI elements and blocks\/items flying to their respective counts were easy to implement. I feel like my biggest win was making item and score counts only increment when the block\/item reached it.<\/p>\n

Music<\/h3>\n

The main riff was based on something I’ve played around with before on guitar–variants of C, with Am and Em thrown in. It worked pretty well, which means I spent about 10 minutes working out the tune, leaving most of Sunday for recording and mixing.<\/p>\n

Unity UI<\/h3>\n

Because I wanted new blocks to fill in like a circular progress bar, I ended up having to mess with Unity UI early (as it supports “filled” images). I’ll confess, it’s something I’ve avoided for the longest time, because it’s intimidating.<\/p>\n

Rather than mess around with the large block of numbers that don’t seem to change anything (Unity just recalculates the X and Y positions when you change them), I’ve preferred to stick to world space, Camera.orthographicSize and Camera.aspect, 2D Toolkit, etc.<\/p>\n

However, I feel like I’ve made serious progress learning Unity UI thanks to this game. And because I didn’t need to dip into 2D Toolkit for text, I think this is the first Ludum Dare where I’ve had no Asset Store dependencies.<\/p>\n

Coding<\/h3>\n

There are few things more satisfying as a coder than being able to call a “Reset” method after a Game Over screen (as opposed to “screw it, I’ll just reload the scene”).<\/p>\n

What didn’t go well<\/h2>\n

Block pop-in effect<\/h3>\n

One of my big plans for visual effects was to have blocks “fill in” like a pie chart. That’s actually pretty complicated if you’re using Unity’s Sprites, and I spent more time than I’d like trying to make it work. Once I realized sprite masking was going to require shader code, I gave up on this approach.<\/p>\n

Even though I was reluctant to mess with Unity UI, UI Image supported this exact feature, and I got the effect I wanted. However, I wish there was a way to do this via Sprites, and I spent more time than I’d have liked chasing that solution. (Although, I did do this first thing on Friday night so I could budget my time.)<\/p>\n

Graphics<\/h3>\n

I’ve been playing around with Krita recently, and because of some of its pen and paint effects, I initially picked it over GIMP. However, the art I created didn’t feel right–it had a dark, hard-contrast feel.<\/p>\n

This wasn’t Krita’s fault, I just wasn’t familiar with it. It was definitely a case of thinking a tool based more on physical mediums would magically produce something “painterly” without me having to understand how it worked. I ended up redoing everything (except the cloud particles) in GIMP, and was pretty satisfied with the result, despite spending a lot of extra time on unused art assets.<\/p>\n

Music<\/h3>\n

Because I guess I have something to prove, I decided to try to mix in cajon, banjo, and mandolin in with guitar. I can’t tell whether it works or whether it’s all out of tune, because my first impression changes each time I listen to it.<\/p>\n

I suspect part of the problem is I changed the speed on the guitar part in Audacity, and then tried to increase the pitch by the same percentage to compensate. Also, I’m not enough of a musician to pull this sort of thing off with consistent quality (which is why you’ll note all of the non-guitar tracks are mixed very low).<\/p>\n

Build and release<\/h3>\n

To be fair, it went OK, but this is the first time I’ve been frustrated with myself for not planning in advance.<\/p>\n

I always go in building for the default Unity 960×600, only to remember on Sunday afternoon that it’s slightly too large for the LD site’s embed (and even a little awkward in my browser). I really need to standardize on a resolution and aspect ratio before I go into a jam. I tend to just jump in the preview window and forget about this part.<\/p>\n

Also, given that WebGL builds seem to take somewhat longer than Unity plugin builds did, I need to start planning for this in advance. Luckily, I did a release to itch.io on Saturday, so I was able to iron out some WebGL issues with particle effect sprites early, but it would’ve been stressful if I had waited until Sunday.<\/p>\n \n

Tags: LD35<\/a>, post-mortem<\/a>, postmortem<\/a>, unity<\/a><\/p>","time":"April 18th, 2016 6:52 pm","title":"LD35 Post-mortem"}