Archive for June, 2010

The Part Of Making Games That Isn’t Making Games

2010/06/24

[Posted to the Gaslamp Games blog.]

  1. Make game
  2. Profit!

Something is missing here, see, and what’s missing is what really does the trick for the commercial indie game development thing. It is those developers that can fill in point number 2 that are successful, I think, regardless of any sort of brilliance in point number 1 (and sometimes making up for a lack of it).

We’re all doing something to carry some of the weight of step two, Derek handling hosting and coding online things, Daniel spearheaded incorporation and is our business guy probably because everyone else hates the idea of doing it more, and Nicholas has shadowy “industry connections”.

As for me? I do art. And it turns out there’s more to a game than the graphics.

(more…)

The Insane Vortex of UI Redesign

2010/06/17

[Posted to the Gaslamp Games blog.]

This wouldn’t be Gaslamp if we didn’t completely redo a major game system once a week.
And this wouldn’t be the ongoing Dungeons of Dredmor beta if we didn’t completely redo between three and five major game systems every week!

Let’s talk about UI redesign.

Here’s the main game UI in Dredmor 0.4:

(Click on any of these images to view at full size.)

Not so bad, right? Rather archaic and clunky, perhaps. But the clunky UI has what we might call character.

I think that the defining feature of Dredmor is not elegant gameplay, great graphics, or cutting-edge technology — it is character: zaniness, a weird ‘take’ on everything. We reference Doom, Diablo, and Ultima with the UI, and I imagine people who understand Dredmor are people who have nostalgia for those old games. And this old thing is finally getting to a useful place with the auto-loading of starting skills and the keyboard hotkey number implemented.

Still, the UI has been problematic and it does look old. I’m quite torn on the issue of revising it, but we’ve had some ideas kicking around and after arguing with Daniel which ended with me coming around to possibly trying a new model for the UI, we started in on The Madness. They key idea is to improve skills interactivity, to make them easy and intuitive to use by having skills act more like items and, at the same time, to have items act a little more rationally. (For example, if you’re standing back and merrily shooting a cluster of enemies with a crossbow and you fail to notice that you’ve run out of bolts, your next click on an enemy will cause you to walk over into the group of enemies … which is exactly where you don’t want to be, presumably because you were off with a crossbow doing crossbow things because you’ve wanted to avoid melee combat.)

A lot of that is about the coded interactivity of the objects. Weird things are happening in the next patch; it takes some getting used to the new metaphor for interacting with skills, objects, and the new, combined quick-use bar, but I think it’s working. And I’m getting ahead of myself.

Back to the visuals: How could we redo the UI to be sleek and efficient, to use the new skills-as-items model? I drew some quick sketches.

I must admit, I actually drew these in reverse order, from 5 to 1, and 1 is based on a very rough sketch that Daniel sent me (by taking a picture of a drawing on paper, of all things). I rather like the top one with the Diablo-style health and mana orbs because it retains the character portrait box and goes with the rather radical move of making major UI elements like the quests, inventory, and skills accessible only through items (which we’d make very, very sure the player could not lose.)

As either a conservative or intermediate step, who knows, I quickly adapted the old UI to the one quickbar model, based on a charming sketch by our dashing lead programmer, Nicholas. [He’ll probably hate me for showing the world his drawing, but I thought it was so adorable that I had to save it … with multiple backups.]

I quite like how the statue on the left has grown a beard in his interpretation.

Here, then, is a mockup of the new minimalistic game UI and then a first draft’s implementation which is in the working code and will likely be released in some form or another with beta version 0.6. This is just the visuals; Interaction is changing a lot as well, but implementation of that is more Daniel’s domain.

It’s quite a change, and it does lose something of the character of the old UI — and the portrait is cut entirely out, but then I didn’t look forward to drawing 28 of those. Hopefully we make up for this loss with some very interesting gains in other areas of the game … which is a subject that shall have to wait to be discussion until another day.

Sketches

2010/06/14

Trying some things. Keeping saturation below 50%.

Darklands is on the mind. I love gothic sallets.

Experimental Perspectives on Tilesetting

2010/06/10

[Originally posted to the Gaslamp Games Blog]

We’ve got a little design problem in Dredmor that Daniel has named “fighting arrows”. See the little arrow at the bottom of the screenshot on the left? It points to a blobby-monster just poking its little eyes out from behind a wall that otherwise covers it up. The arrow is a helper icon to make sure you notice that there’s a monster.

No, this is not elegant. We’ve also got issues with doors being difficult to see behind walls. Well then, how do games deal with the problem of stuff hiding behind walls?

One solution which came up was that of Zelda: A Link to the Past — they made it so that there is no ‘behind’ walls. See the right screenshot: everything has a rather subjective take on perspective. The player sees the face of all of the walls, no matter what direction they face! One column is seen from the front, another is seen from the right, and there is even some weird overlapping balcony thing. The world of A Link to the Past has a take on perspective that would make Escher proud, and the game manages to get away with it.

Could we?

I threw together some test graphics last night.

It would be painful to redraw all the tilesets, but it could be done, and a couple could be cheesily re-colored or something to cover all the dungeon levels. But what it comes down to is that the perspective our player character and monsters are drawn in is way too much looking like a strong front-view. All the characters in Zelda are, if you notice, rather squat, like they’re being viewed largely from above. And they’re smaller, just a bit more stylized, visually contextualized in a looser manner that lets them exist in an environment with a fantastic interpretation of perspective.

From talking it over with Nicholas, it sound like we could have gotten away with it if we’d designed the game this way from the very beginning (which was what, six years ago? And I’ve only been around for nearly two. Oh, and how many things I would change if I could have been there to help design the visuals of this game from the start!) Alas, we’re locked into a certain direction with only so much room to maneuver. Re-writing the rendering and dungeon-building code and redrawing all the tilesets is a bit too extreme a maneuver, I’m sorry to say. And before anyone suggests it, I think Nicholas would go into some kind of coffee-fueled berserker rage if we suggested he re-write the rendering code to support transparency masks or wall opacity when we’re halfway(?) through the game’s beta.

Don’t worry, we’ve still got tricks up our wizard’s sleeve.

The Interaction Problem

2010/06/03

[Originally posted to the Gaslamp Games blog.]

Oh playtesting, how you tear down my illusions, besiege the fortress of my ego then poison its well and set fire to its stores of grain.

Fig. 1: The good part of Dredmor’s interactivity.

It is shocking just how surely a player will ignore tutorial text. The help button is effectively invisible, ignored, the text left sad and unread. Whatever it is, the “go away” button is clicked via Skinnerian response to years of training at ignoring inane popups. Yes, Nicholas passed me a link (or possibly a newer one) to Jeff Atwood writing on the subject when this issue of the tutorials being completely ignored came up, and it got me thinking.

Fig 2. The bad. Don’t ask why it says “Axe” on that lightbulb.

We’ve got issues

1: We need some kind of hook to get people to notice the tutorials, some way of breaking them out of the usual ignore-tutorials behaviour. I’m thinking of replacing the starting text dialog pop-up with an informative one-panel mini-comic which points to the tutorial button and the skill bar, which should be what people absolutely pay attention to. Everyone love comics, right? Or maybe it’s that people really hate lightbulbs.

2. We need to make our interface more intuitive, or at least cover the intuitive expectations a player makes of our UI; A player should not really need to read the tutorial if they’re familiar with common UI schemes of games in and around this genre.

3. Absolutely no one has understood using the skill tome and skill bar without being explicitly told how to do so. You have to open the skill tome with the skill button at the bottom of the screen, then drag appropriate skill icon to the skill bar at the lower right of the screen. Then you must left click to select a skill in the bar, and with a skill selected right click to use the skill. Possibly on an appropriate target.

Fig 3. The ugly.

Actually, that whole process of readying skills to use sounds confusing when I write it out. I don’t think we noticed lots of things like this because we’ve been trapped in a bubble staring at our work for over a year. (So this is how MOO3 happened! That and a pushy publisher, which is a problem we’re lucky enough not to have. Ahem.)

On the other hand, there are interactivity schemes I’ve noticed people take to almost immediately. Pointing and clicking to move around and use things is something everyone seems to know. The first thing players do is start walking around and clicking on vases to break them, on enemies to attack them, then on items to get them – and when the item is attached to the cursor, it is intuitively dropped into a slot in the inventory bar.

Let’s look at Fig.1 again:

I notice now that we’ve been mixing metaphors with game object interaction: Why do skills not work like items? Everyone understands how items work in Dredmor with little trouble, can we not let a player ‘pick up’ and use skill icon instances the same way we let them use items? This demands baseless conjecture and rash experimentation. I have notified the relevant authorities.

Also:

Fig 4. The weird.

No one at all has asked me about this little section of the interface and I don’t really know why; From all I can tell no one has even noticed it exists.

UI design boggles the mind.