I'd say that you should take your average, nothing-special platformer, and then remove something from it. A platformer I'm working on took out the ability to control character movement; he just runs back and forth into walls and all you can make him do is jump. Just those small changes can make it feel different.

Or like the idea I had once: a game where you can't jump, so traversing the y-axis becomes some sort of a puzzle. Maybe you have to collect one-use items that allow you to jump, then use them strategically. If the items are permanent entities that remain on the map, you get further gameplay depth, as they can help or hinder depending on the placement.

For me, it's that core mechanic that comes first, and only then do I start working on the engine. One of my strengths is knowing how a game should "feel" like right from the start, so it's hard for me to imagine doing graphics or gameplay before having the direction pinned down.

And of course, you can expand the game's world from there - what sort of character can't jump? What does the character think about lacking an ability? Like, maybe you're a wheelchair patient out to prove something, and the items/puzzles/entities you encounter are hospital-themed. Or it could simply be an abstract, arcade-style game with grid-based movement. It often depends on what games you've been into recently.

I wanted to create... a happy place...

Even my imagination isn't pure anymore...

awww yeaaa

It's more than a game to me, it's like therapy. The baroque graphics and music, the otherworldly setting, tons of tweaking and relaxed grinding, with a plot that goes beyond "hero defeats evil empire"... all that makes me feel happy and inspired.

As I've become more absorbed in gamedev, I've started to appreciate games in a different, deconstructive way. I spend most of the time thinking how it's done and how it could be done better.

Like, "this sprite was really drawn by someone like me", or "someone like me spent an afternoon coding this event". Before, I just tried to throw myself into the experience and spent less effort on analyzing the implementations. Playing games and reviewing what's successful/flawed in them is a major source of inspiration.

On the flip side of the coin, games have become more disposable for me. There are still many that catch my attention or even get me addicted, but once you learn to see the patterns in game design, you end up realizing that so many games are the same thing with different presentation. Kill enemies, get items, move on.

So you've read through programming books & tutorials. You've become quite efficient at using control structures, data structures, functions, and object-orientation by applying your knowledge in small projects.

But where to go for larger design patterns? As in, designing your project so that it's scalable, utilizing programming interfaces and management classes, maximizing performance... There seems to be relatively scarce information about this sort of thing out there, as it's not really covered in programming tutorials.

The book Code Complete is one good resource for software development in general, but is there anything gamedev-specific - whether it's articles or other documentation?

Never7 isn't nearly as good as Ever17 or Remember11.

Yeah, and I don't think anyone was expecting it to be. But it's got the same charm as listening to forgotten releases by kickass bands, back from when they were just a couple of guys jamming in a garage.

Still need to read that some day. Apparently it's a classic.

Get ready for a lot of "holy fuck just tell me where I'm supposed to click ararararagh". Sometimes the logic behind in-game events occuring seems random, and I end up walking back and forth between places like a Pokemon trainer in tall grass. Or maybe I'm just bad at point'n'clicks. And many of the eroge scenes seem out of place.

Otherwise the story and presentation are neat, it's leaps ahead of similar games from its time. The "jewel save" gameplay mechanic that YU-NO revolves around is pure brilliance.

It's getting the same treatment as we speak

Spoke too soon then. I'm glad to have been wrong about this one.

[Edit]

Daily reminder that you will never experience Coco's route for the first time again.

For some less depressing reminders: the first Infinity series VN, Never7, was recently translated. As was the oldschool game YU-NO, which is like a crossover between point'n'click and VN that Infinity fans will surely love. Now if only I/O (from the mastermind behind Infinity) ever got the same treatment...

Oh, and 999 has a sequel called Virtue's Last Reward, of which an official translation is scheduled for next month.

Edit: actually October 23rd in North America. You know it's going to blow your mind.

Oh wow, this looks really immersive judging by that video. Time and time again, I'm blown away by the gamedev talent up here in the north.

http://jeremykun.wordpress.com/2012/07/29/the-cellular-automaton-method-for-cave-generation/

Also, ANSI Common Lisp by Graham for a CS class. Lisp is pretty fuckin' cool.

Practical Common Lisp for me. why-can't-i-hold-all-these-parentheses.lisp

I'm bothered by the slightly stretched graphics, you can see resolution mismatches in the backgrounds due to upscaling.

A sight for sore eyes. The games look delicious and the added graphics are tasteful enough to enhance the collage without taking the spotlight.

Well, this has gone places from last I checked. I'm still not a fan of the graphical style though.

i think we at tigsource should make a show like this, featuring different developers and how they take the approach to designing games. i think that'd be a much more valuable show than most of these, because we actually make games, and most of these people don't seem to

I support this idea. Great advice often pops up in design discussion threads, and it'd be useful to condense our collective knowledge into a (sort of) educational format.

I'm going for a 100% completion in Bio Menace on Hard.

That's how this game is supposed to be played. Every HP counts, you have to conserve ammo and consider strategies, as if you were playing a roguelike. Each try brings you closer to clearing the level, and when you do, it feels rewarding.

Right now I'm stuck on the Queen Ant boss in Episode 2, because I wasted my lives in the previous secret level, and there are only two machine gun clips in the boss fight (plus grenades which take quick reflexes to get). I already came so close once, there were only like 4 pixels of its health bar left, and I was cussing at my screen.

I kind of wish I'd recorded a Let's Play, this is hardcore.

It looks great, yeah. I think there's a lot you can do with that Aztec theme, not just graphics- but also gameplay-wise.

Inventor: Doom
Master: The Half-Life series
Diluter: brown WWII shooters

Down-voting isn’t even really helpful in recognizing the diamonds from the rough. If the goal behind Greenlight is to find the games that most people will buy, then down-votes are utterly unimportant; it’s not like that translates to a “negative purchase” should the game come to Steam. Think of this: if a game gets 1 million upvotes and 2 million downvotes, it would have an average of only 33%, and negative one million votes–even though, in reality, a game like that would be a highly profitable release with a million purchases! What aspect of a game’s success on Steam is the down-vote portion of Greenlight even trying to represent?

That's true.

But despite some implementation flaws, I like the idea and am disappointed by the cynicism.

Tower of the Gorillion looks promising, I had an idea for a game like this once.

Since technical limitations don't dictate sprite dimensions anymore, you're free from worrying about emulating old styles.

It doesn't have to be a square, it doesn't have to be a power of two. Spend your energy on making the character look like it does in your head.