Good Robot #37: Get Things Moving

By Shamus
on Nov 4, 2015
Filed under:
Good Robot

Let’s talk about little things that make a big difference.

The game was lacking something. It was just too static. The world didn’t move, didn’t animate, didn’t react. The screenshots looked good, but when you were playing the game it felt like you were flying past a painting or something.

So my first solution is dust particles. Here is a shot of them, zoomed in so everything looks terrible:

Good Robot 2 will feature a multi-pass hypo-allergenic dust diffusion and anisotropic pollen filter, but only on the latest NIVIDIA cards.

Good Robot 2 will feature a multi-pass hypo-allergenic dust diffusion and anisotropic pollen filter, but only on the latest NIVIDIA cards.

The little specks are the thing we’re interested in here. In still frame, they just look like part of the background. But in gameplay they drift around. They get brighter when you hit them with your flashlight beam and they’re obscured by walls.

Like a lot of “field” effects, this one is pretty simple. There are only 400 dust motes in the game. They drift slowly, so the game doesn’t even have to move all 400 every frame. Each one of them moves in a different (but constant) direction. If it drifts off one side of the screen (either because you’re moving, or it is) it simply re-enters from the opposite side. If you could zoom out, you’d see your character has a rectangle of white particles following them around.

It sounds simple, but it’s amazing how much more alive this made the game seem, particularly since the movement is so slow.

So I thought that maybe we’d get an even bigger return if we took it to the next level. I made a version where the dust particles would tumble away from you in your wake. The result was… nothing. I barely noticed the effect. It also took more than twice the CPU. So I’d managed to hit the sweet spot on the cost / benefit curve on the first try, and then took it a step further and realized I’d wasted my time. (I undid these later changes.)

The other big thing I added to the game is the concept of machines. I mentioned them earlier in this series, but skimmed over the details.

Warranty

The game is creeping towards rogue-like gameplay as it develops. We’ve added permadeath. However, you can sometimes find a vending machine that will sell you a warranty. If you’ve got a warranty, then when you die this machine will put you back together:

I know the extended warranty is usually a rip-off, but trust me. THIS ONE IS WORTH IT.

I know the extended warranty is usually a rip-off, but trust me. THIS ONE IS WORTH IT.

Being revived consumes the warranty. If you can reach another vending machine you can buy another warranty, but they get more expensive every time. That will consume precious money better spent on upgrades. I like this because it forms a nice compromise between the severe cruelty of a roguelike and the coddling of an action game. Making a single mistake won’t end your whole game (Spelunky was actually way too frustrating for me) but you will still carry the burden of that mistake going forward. The money spent on the warranty means you’ll be getting upgrades just a little later. (Or if you keep dying, way later. Or never.)

If this kind of roguelike gameplay sounds a little strange coming from me, it’s probably because I’m part of a team now. A lot of these gameplay decisions emerge after literally hours of debate and playtesting. Everyone has their favorite gameplay mechanics and styles, and a lot of the current design process is looking for a mix that makes sense to everyone. There have been a few decisions that I disliked as a matter of personal taste, but when other people seem to be really passionate and enthusiastic about an idea, I’m willing to let it go. If they love it more than I dislike it, then it’s probably a net win. As long as the final whole makes sense, then it’s all good.

There’s always a trade-off here:

Danger #1: Compromise too much, and it might ruin your creative vision. You won’t end up with the game you wanted to make. If the game does poorly, you’ll always worry that maybe you should have fought a little harder for that thing that might have fixed it.

Danger #2: Never compromise, and you won’t have any way to detect and correct bad decisions. David Cage, Peter Molyneux, and the writers of the Mass Effect 3 ending are all people that seemed like they need to be more responsive to their teams and self-aware about their excesses. You’ve got smart people on your team. Trust them!

Since this will be the first time I actually ship a game, I’m more worried about danger #2 than danger #1. If I design a hit game, then maybe I can think about embracing the mentality of a swaggering, uncompromising auteur.

Anyway. Back to the machines.

Machines

Purple? I APPROVE. YOU CAN DO ANYTHING YOU WANT WITH THE GAMEPLAY. JUST MAKE MORE THINGS THAT ARE PURPLE.

Purple? I APPROVE. YOU CAN DO ANYTHING YOU WANT WITH THE GAMEPLAY. JUST MAKE MORE THINGS THAT ARE PURPLE.

The goal wasn’t just to put some detail on the bare rock walls, but also to add some stuff that was animated in some way. The little gears in that respawn machine actually rotate, the pistons on top pump up and down opposite one another, and some of the lights blink on and off at fixed intervals. Like many other things, machines are defined in text files like so:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
<Spawner Message="" Use="spawn" Offset="0 0.4"  MaxActiveRobots="6">
    <!-- Main part of the machines -->
    <Part0
      X="0"
      Y="-0.050"
      Depth="0"
      Color="#FFF"
      Glow="0"
      Size="2.5"
      Spin="0"
      Sprite="machine9" />
    <!-- piston 1 -->
    <Part1
      X="-0.66"
      Y="-0.50"
      Depth="-0.015"
      Color="#fff"
      Glow="0"
      Size="0.45"
      Spin="0"
      Sprite="machine6"
      Move="0 -0.12"
      MoveTime="-2000"
      MoveLoop="1" />
    <!-- edited out a bunch of parts so this would fit in this blog post -->
     <Part11
      X="0.8"
      Y="-0.055"
      Depth="-0.05"
      Color="#000"
      Glow="0"
      Size="0.25"
      Spin="0.66"
      Sprite="machine5" />
  </Spawner>

The artist can specify any number of parts of the machine. They can give it a specific sprite image, offset from the center of the machine, rotation, and color. They can also tell it to spin, or move back and forth between two points.

This gives us some moving, blinking, spinning crap to look at, which saves the game from looking too static.

Enjoyed this post? Please share!


202020203There are now 83 comments. Almost a hundred!

From the Archives:

  1. Galad says:

    Roguelike gameplay? Permadeath not being the permanency it usually is? YES! *throws money at screen*

  2. Da Mage says:

    I know this question is ancient history to you know, but what did you do when it came to your GUI? Did you go with an existing library or did you make something yourself.

    Anyways, looking great, can’t wait until it finally releases.

  3. gstnk says:

    I’m sure you said previously that you wouldn’t use XML. Maybe I’m misremembering? Anyway, that example is surprisingly readable. Is it all that clean or has it gone the way most uses of XML seem to go?

    • Primogenitor says:

      The previous post in the series was “my custom data file reader is too slow”.

    • Shamus says:

      It depends on the file. Arvind dropped in an XML reader when Pyro joined the project. Simple data is in ini files, and complex data is in XML.

      The machine files are pretty readable, although they do get kind of verbose. Just one machine takes several pages. It’s probably easier if you’ve got an editor that can collapse and hide nodes. Mine doesn’t, but I hope the art team does, since they use these files all the time.

      • Matthias says:

        Have you considered using Visual Studio (which you’re already using for coding if I remember correctly) to edit XML files? It’s what I use at work and it’s actually really good. If you write an XML Schema Definition (XSD) you can even get autocomplete, documentation and validation.

        • King Marth says:

          Agreed, I’m not sure what editor you’re using for the XML but just so you know Visual Studio should really have those features you seem to be missing out on.

          Another good option is notepad++, which should pick up on a .xml extension (or you could set the language directly) and allow you to at least collapse nodes with a corresponding close tag; I tossed your sample into a new file and it doesn’t do much to help with multi-line tags full of attributes. It did allow collapsing the spawner tag enclosing the full thing though, so it’s better than nothing.

          • Blue_Pie_Ninja says:

            there is a plug-in for notepad++ that does this, I believe it comes standard or you have to install it but I used it when I was modding Saint’s Row 3. The only problem is having to manually insert line breaks to properly read the damn thing as well as tabs but it was worthwhile as the collapsing feature worked fine.

            • WJS says:

              You put line breaks in manually? *Hysterical laughter*

              …Seriously, what? Never mind using an existing prettifier, it’d be quicker to write your own!

              (If all you were doing is line breaks, not indenting, sed would do…)

      • AileTheAlien says:

        For your next project, I recommend the YAML format. Then you get the not-heavy-with-brackets-everywhere syntax* like an INI, but with the flexibility of an XML file. (I had an example here, but couldn’t figure out the tags to make it show up with indentation.) Just look at the examples on Wikipedia that have hardly any brackets or punctuation at all. So clean! :)

        * Originally, only non-bracket-filled style was allowed in YAML. Later, the syntax of JSON was added as part of YAML, so you could make stuff that looks gross (to a human), like this: {“robot1”: {“weapon1”: “missiles1”, “weapon2”: “lasers3”}}

      • Naota says:

        Interesting side-point: I actually prefer the .ini files to XML, and I’m the guy who did the overwhelming majority of the scripts in Unrest (and the updated version of WFFF) using the latter. I’d say this has something to do with the fact that I started out scripting and design-monkeying early 2000’s shooters, but it’s more likely that XML is incredibly verbose and just not that elegant to work with.

        That said, the way it’s used in Good Robot is pretty trim and straightforward, so it’s never really bothered me.

  4. MichaelGC says:

    sell you a warranty

    I smell microtransactions! And I see everyone with an AIMD graphics card is getting screwed-over again. C’mon, you might as well announce the various confusing tiers of Day One DLC now and get it over with.

    • krellen says:

      I have a job because of extended warranties. I am not allowed to denigrate them.

      Just yesterday I fixed a problem by plugging in a cable for someone. Some people really should get the warranty.

      • Lanthanide says:

        Actually there’s some scope for some funny feature creep!

        Make the default Warranty only bring you back to 50% health or something. And you have to buy an Extended Warranty to get you up to 100%.

        Unsure if this extra concept would be worth the potential confusion / interface additions, though.

      • Peter H. Coffin says:

        The world should really probably call them “service contracts” and be done with it. “Warranty” in my mind (and probably to a lot of others) implies that dissatisfaction with performance or function (it doesn’t work with *my* computer even though it may work just fine with someone else’s) is enough to get a return and refund, while “extended warranty” almost always limits to “we’ll fix it if it breaks in a way that’s inconsistent with expected wear that happened with expected use”.

        • krellen says:

          Given the things I fix, I’m not sure my company agrees with your definition of extended warranty. I don’t fix damage from dropped machines (unless they bought that coverage too), but I do fix a lot of problems that are basically regular wear-and-tear issues.

  5. Zak McKracken says:

    So … I guess the missing link between the last article and this one is that you solved the slow data loader problem by moving to XML after all?

    Also, I’m probably a bit too old-school to appreciate permadeath… If there’s one bit I can’t quite get past, I very much like the ability to try only that bit again without restarting from the very beginning. Most games back in the day would give you x lives plus 1 per level or somesuch, so when you make it all the way to the end but die to the final boss you can just try the final boss again. In the last twenty years, my tolerance for that situation has not increased, so I’d be in favor of maybe having a game mode that isn’t as unforgiving..?

    • Ingvar says:

      I don’t know about “old-school”…

      Permadeath is strongly associated with “Rogue-like”, which harks back to 1980. That’s actually later than I started playing computer games, but not by very much.

      I guess the more classic model is “limited, low, number of lives” as made popular with arcade machines (sometimes augmented with “push more cash in for a continue”). This ended up in (some) non-arcade games, for consoles and computers and whats-have-you, occasionally with a limited number of continues.

      I’d say that I don’t mind either way, it depends too strongly on the game.

      • Joe Informatico says:

        We’re both old enough to remember early 80s gaming–both arcade and home–where the concept of lives and continues was the base assumption of, well, almost the whole medium. To see Shamus go through this detailed in-world explanation of basically the same concept in Good Robot makes me feel really old. I don’t know, it’s like explaining to someone who’s only ever driven cars with automatic transmissions that there was once a time where all cars had a manual transmission. (This metaphor works better for North Americans, I guess.)

        • Ingvar says:

          In at least the UK and Sweden, if you pass your license test in a car with automatic transmission, you’re not licensed to drive manual-transmission cars. The latter is still a large proportion of the cars on the roads.

        • Ivellius says:

          Oh. I hadn’t–

          But when you put it like that, that’s exactly what he’s described

          and it makes me feel old too.

      • Zak McKracken says:

        All the 80’s games I have in mind* have a limited number of lives, and most will give you more as you progress, but I don’t remember any where you had to decide between upgrades and lives. Typically you’d just lose some upgrades when you die and have to work to get them back. Even that was sometimes frustrating because the second try for any challenge you fail is harder than the first. If you need to buy lives from the same resource you’d otherwise use on upgrades, failing makes the entire rest of the game harder, not just the next bit.

        * Mario, Giana Sisters, Boulder Dash, Space Taxi, Pac-Man, Pitfall, Moonbuggy, Katakis (almost a 90’s game), Last Ninja II …that thing where you bounce a tennis ball over a maze… the first time I came across the term “Rogue-like” was maybe five years ago, referring to Indie games … I quite possibly grew up in a slightly different world?

        • Nidokoenig says:

          The general idea behind taking away upgrades is to force you to up your game. Take the example of Contra, you lose something that spreads, or does more damage, or rapid fires, and you instantly know you need to be more careful, aim better, dodge better and generally last longer and be more accurate. This puts you in the mind set to learn enemy patterns and focuses your attention, which builds towards the “perfect” run where you don’t die. This is for games that last less than an hour on a good run, obviously.

      • Blake says:

        Coins for lives you say?
        I wonder how long it will be before publishers start re-releasing old games with microtransaction lives.

        SPEND FUN-DOLLARS TO CONTINUE

  6. Ringwraith says:

    No wonder Shamus liked Saints Row so much, their colour scheme is purple!

  7. MrGuy says:

    So…you added an effect that looks like crap in a screenshot but good in a video and gave us a screenshot of it? Moar vidz plees!

    Seriously, I always get fired up to buy this every time you put another video of the gameplay up. I love that little guy!

    • MadTinkerer says:

      “an effect that looks like crap in a screenshot but good in a video”

      Shhhh! Don’t let the marketing department at Pyrodactyl know this is happening! Shamus will be forced to submit all of his blog posts to PR for approval!

      • Naota says:

        Worse: I’ll have to record more gameplay footage.

        “But Ross!” you say. “You run an LP series that’s done everything from FEAR 3 to Twinsen’s Odyssey. How could you possibly have recording troubles with Good Robot?”

        Or you would, if my LP series had more than a handful of viewers at a time. Well… as succinctly as possible: remember that time Josh had to repair his sink?

        The expected resolution for Good Robot’s media is 1920×1080, and I have a 16:10 monitor that clocks in at the just-barely-not-enough 1680×1050. It’s a pretty good monitor, and not in any immediate need of replacing… and every other member of my family also has a 16:10 monitor.

        “So then how-?”

        I have to get an up to date build of the game on my laptop, hook it up to my television, so I can run at the correct resolution. I would record now, but first I have to connect an external HDD – my laptop only has 10 GB free and this is uncompressed video, you see. Finally, Good Robot is in a position to capture video. At exactly half its native FPS. Due to an OpenGL bug with video capture. So… I have to play in slow motion, and double the playback speed of all of my clips when I edit the video.

        Just in case you’re wondering what sort of shenanigans go on under the hood of a simple video file – a lot of them. All the time. I’ve become a practiced shenaniganmancer.

  8. Zoe M. says:

    Can you make destructible bits on the walls? Static elements like thin spikes or abandoned machinery that you can destroy for a small payoff in money?
    That’s something that in my experience makes the game world more engaging – it shows up in lots of older platformers as something to do aside from just killing enemies – and it would also make things feel less static.

    • John says:

      It’s also in Bastion. Of course, in Bastion (i) the payoffs are indeed small and (ii) there’s a narrator who comments on your actions. “Kid just rages for a while.” I always felt kind of silly stopping to destroy absolutely everything when I was supposed to be busy undoing the end of the world.

    • Khizan says:

      I hate this kind of thing because it starts to just be boring bookkeeping crap you have to do to keep on top of the gold curve. First you kill the enemies, then you smash all the stupid pots, and it begins to just be a boring hassle.

      I hate this most of all in roguelike variants because you never know when that 3 bucks you passed up will be the difference between life or death. You got tired of it and didn’t want to smash every wall to check for destructibles so you skipped some and now you’re 5 bucks away from the supergun and screwed.

      • AileTheAlien says:

        I think it could work if the scenery-smashing only gave small amounts of money, and upgrades ramped up in price. Then the player gets a bonus early in the game, if they are patient, and later when they’re bored, they can safely ignore it. Alternatively, just let players smash terrain for fun, with no in-game incentive whatsoever. Fun smashy times! :D

  9. Bitterpark says:

    Check out this talk on making games “juicy” and responsive to the player if you haven’t already. It was one of those things I never really noticed in games, but couldn’t stop noticing after watching that video.

    • krellen says:

      I first saw this when Shamus linked it (probably on the blog), so I’m pretty sure he’s seen it.

      Still good.

      • Bitterpark says:

        Well, silly me. That’s probably how I found it in the first place.

        Still, the post was addressed to whoever reads it, not just Shamus.

        If you (yes YOU, whoever is reading this right now) are at all interested in game development and somehow missed this, you should watch it. It teaches an important lesson: a game should react to player input in exciting ways: pressed buttons should depress and make sounds, hit objects should emit particles to give the player an impression of impact and so on. It’s an important part of the feedback loop.

        • WJS says:

          That’s not just a gaming thing; feedback channels is incredibly important in just about any interface. Take a cockpit for example; despite advances in technology, fancy HUDs and the like, they still use buttons and switches rather than touchscreens. A pilot needs to feel that he’s operated a control, not just hope so. Haptic feedback is incredibly important.

    • Abnaxis says:

      Wait…

      So I watched that video the first time I saw it reference on 20 sided (don’t remember if it was Shamus who posted it, but I definitely saw it here), but I decided to watch it again.

      Long-hair guy (Petri, I think? Didn’t catch their introductions..) sounds a lot like Jarenth to my American ears.

  10. Orillion says:

    I don’t know, kind of sounds like your team is nudging the project in the direction of something the market is currently saturated with. That’s the third danger when designing with a team: “I liked X game and A mechanic so let’s see what we can do with that” ignoring the fact that Y, Z, V, H, W, Q, S, T, and P were all games that had that mechanic and all were released in the past two years or so.

    I may be biased, though; roguelikes just don’t appeal to me and roguelikelikes are somehow worse

    • Bitterpark says:

      I love both and I can’t get enough.

      I don’t give a crap about shoot-em-ups, so I wasn’t going to buy Good Robot (sorry Shamus), but if it has roguelike elements I’m in!

      At this point, I’m not sure if roguelite is even a genre. I think there’s plenty of potential for combining existing genres with perdmadeath and procedural content generation, and this new wave of indie games has only just started tapping it.

    • Nidokoenig says:

      I’ve wondered before about adding permadeath and other challenges as an additional challenge you can select, since if you build it to make permadeath acceptable for, say the top five to ten percent of players, you’re forced to cut out a lot of nonsense like gotcha moments and forced damage(beyond some environmental if they want to take a shortcut without getting the immunity upgrade). Obviously, this is more suited to games were you might get beaten down to a game over once every ten-sixty minutes if you’re playing less than optimally than a high speed twitchy game where it could happen in less than a minute.

  11. Mephane says:

    Since a lot of the game’s content and assets will be defined in modder-friendly text files, does this apply to the warranty machines, too? In particular, as someone who utterly hates permadeath mechanics, I would like to mod the machines so that the price never escalates, and that the machine can be reused (i.e. you are not required to find a new machine it get another warranty, i.e. you are never in a permadeath situation unless you run out of money). This may sound like cheating to proponents of roguelike game mechanics, but is actually a staple thing in other games.

    For example, when you die in Borderlands 2 you lose a certain sum of money and respawn at the nearest teleporter station. It took 7% of your total money, so technically it is never game over, when you are completely broke, 7% of 0 credits is still 0. This had the interesting side-effect that the cost of death scales up with your progress in the game (reflected in your money), but scales down if you are doing badly, and it encourages you to actually spend the money instead of hoarding it forever.

    Actually, thinking about it, I would love to see this precise death mechanic to appear in Good Robot. Could that be made as a set of option in the text file? For example, the default could be

    MultiplyEachDeath=2.0
    Penalty=100.0
    PenaltyUnit=$

    Which I might mod into

    MultiplyEachDeath=1.0
    Penalty=7.0
    PenaltyUnit=%

    • Daemian Lucifer says:

      I liked the death mechanics in shovel night the most.When you die,you drop some of your money(25%).If you reach that point again,you can gather it all back.But if you die before reaching that point,you drop some of your current money,and only then does the money at the place where you previously died get lost forever.So you can die any number of times without actually losing any money as long as every time you manage to collect the bags youve dropped before dying again.

      • Mephane says:

        That sounds really interesting. Also gives an incentive to revisit the location and not just abandon a difficult area after the first death. But the principle is still the same, you lose a percentage of your money and thus can never truly become “too broke to respawn”. Which is what I like about the system.

  12. ooli says:

    permadeath with save system?
    I’m conflicted about that.
    Sound good on paper, but there must be a reason nobody in hundred or more of very elaborate and diverse Roguelike never did it.
    It’s kind like the worse of 2 worlds: wont appeal to RL fan, wont appeal to more relaxed player.

    Ask Josh, but if dying under warranty have you get back a few minute of game behind, you’ll probably never bother with it. It’s frustrating enough to die, but there something so awful about having to redo the same part of a level…
    Consider having just a second life on the spot if you die and are under warranty.. Or just a normal free save system.
    Cause optimisator like Josh and I, will never use sweet gold to buy a warranty so awful when you can spend it on nice bigger gun.
    And casual like you, will hate a game were you have to redo some part of the same level.

    • guy says:

      Objection!

      There’s a space roguelike called Transcendence that includes almost exactly this system, letting you buy late-game insurance at huge prices, though they rise each time and after the third no one is willing to sell you a policy at any price.

      You can also restart at the last stargate you jumped through for a huge score penalty.

      • ooli says:

        Is the mechanic fun to play?

        Are the level re-generated once you’re spat out of the stargate? Can you exploit the system to get better loot/score?

        Did people use the feature, or the money/score is always better spend elsewhere?

        • guy says:

          It’s fun to play.

          The levels aren’t regenerated, but the score penalty is so steep that’s not a useful exploit.

          Some people used the feature, other people didn’t, or just used the insurance and not the gate respawn.

    • Nidokoenig says:

      Actually, I’m pretty sure being able to get one auto-revive or extra life isn’t unheard of, it’s just usually a late game or deep side quest thing.

      One thing I’m wondering is how it interacts with the score system, presumably losing a life should impact your score, otherwise a late game section could be farmed until it stops paying for itself. Buying a warranty could also have some effect, so people who want to maximise their score would have to play without a safety net. Or maybe have warranty stations be destructible for money or points, kind of like Shovel Knight.

    • Cybron says:

      I feel like the odds that you’ll NEVER buy one is overstated here. Even speedrunners (compulsive optimizers at their finest) tend to do things like ‘safety saves’ when appropriate.

      Think of it as adding decisions to gameplay.

    • Steve C says:

      Maximo: Ghosts to Glory had a increasing cost to death. Players also had to buy saves too. Saving the game was a limited resource .

      I don’t know how well it worked at making the game fun. I’m not a fan of roguelikes. I would always cheat and turn off the game instead. That didn’t increase the cost of dying.

  13. Warclam says:

    Purple! YESSSSS! I’d play a roguelike if I exploded into beautiful purple on death.

  14. tzeneth says:

    How do you tell when you’ve crossed the line from too static to having a too busy background that can distract from what’s really important?

  15. Cybron says:

    I was not super interested in the idea of this game at first but the drift towards rogue-lite gameplay is winning me over. So yes, trusting your team seems to be working from where I’m sitting!

  16. Ilseroth says:

    Honestly, considering your key issue was pushing the game from “bubblegum game” to something a little meatier, the roguelike path is one of the most reasonable means to do it.

    However this does have a bit of a switch, as you no longer expect a person to take a linear path through your game. Instead, you are getting to a point, dying, getting a bit further, dying and so on and so forth, which is perfectly acceptable… but it requires developing a game with semi random gameplay. I know you guys were talking about lots of guns and whatnot, so I suppose that acts as the primary differentiation between runs.

  17. Chris says:

    You’ve probably answered this elsewhere, but will there be levels of difficulty to the game? Something like an easy-mode where killing a boss will net you a free warranty (or money if you’re already covered), which would enable the folks who want to avoid the perma-death specter a little bit longer..

    Is there any chance at buying an actual CD-copy of the game when it comes out? That’d make giving it as a gift to friends without Steam/internet a whole lot easier..

    Also, whens the beta coming out? :D

  18. Shamus, why not allow people to quicksave, or have other options for saving that can be enabled or disabled at startup?

    On a small project like this I completely understand that creating a save-game system and debugging it is non-trivial work, but this does seem like the Objectively Right Thing To Do from a design standpoint.

  19. NotSteve says:

    So is the next game going to be all about pushing purple people off of purple ledges? (Game of the year!)

  20. Neko says:

    The warranty thing sounds a bit like the Ankh in Spelunky – a one-time resurrect. Typically, I find you die to something stupid the instant you get the Ankh.

  21. RCN says:

    It is so strange to read Shamus openly drooling at the color purple (and part of the SW crew).

    It is a cultural thing, I think. I mean, I LOVE purple, it is basically my favorite color, but here in Brazil purple is considered prominently a “gay” color. And, while I really don’t care about how I’m being labelled by other people, there’s certainly a bit of unease and embarrassment at the thought of other people associating me with purple. Heck, it is probably worse than PINK.

    So… weird. But to hell with it, purple is awesome. Such an energetic color.

  22. The Mich says:

    I’m really looking forward to playing this :)

  23. psivamp says:

    Finally added Good Robot to my Steam wishlist.

  24. Phantos says:

    (Spelunky was actually way too frustrating for me)

    Oh man. I’m getting red in the face just thinking about Spelunky.

    “Hey, let’s make a game where you need 10 hit points and 20 bombs… and then give you less than half of that… and then make everything cost too much money. And make it so when you do have enough money, no shops spawn. And when you die, you have to start all over because what are ‘checkpoints’?”

    People should study Spelunky on how to not make a video game.

  25. Yadu says:

    Is that xml handcrafted? If so, have you thought of a small editor of sorts for the artist drag drop and create -> then generate xml and tweak?

  26. Rick says:

    Sounds great and I can’t wait to play it even though I usually hate punishing games.

    Nobody mentioned the base 2 numbering of machine parts yet?

  27. Roxor says:

    Permadeath and respawn machines with an increasing cost to respawn?

    Sounds like a convoluted way of implementing the classic “you start with three lives and it’s game over when they’re all used up” mechanic.

Leave a Reply

Comments are moderated and may not be posted immediately. Required fields are marked *

*
*

Thanks for joining the discussion. Be nice, don't post angry, and enjoy yourself. This is supposed to be fun.

You can enclose spoilers in <strike> tags like so:
<strike>Darth Vader is Luke's father!</strike>

You can make things italics like this:
Can you imagine having Darth Vader as your <i>father</i>?

You can make things bold like this:
I'm <b>very</b> glad Darth Vader isn't my father.

You can make links like this:
I'm reading about <a href="http://en.wikipedia.org/wiki/Darth_Vader">Darth Vader</a> on Wikipedia!

You can quote someone like this:
Darth Vader said <blockquote>Luke, I am your father.</blockquote>