Hunger Games, built live with Pallarium πΉ
One forest. One way out. Every shot is real gameplay from the game I built in the browser.
#gamedev#indiedev#threejs#AI
@NbhdVideo a horror VHS in a video store is such a good container for clues, because the tape itself can lie to you. do the clues ever change on a rewatch, or is every tape fixed once you find it?
Moscow, 1908 to now. snow on the Kremlin, lights over Red Square. βοΈπ·πΊ real archive footage, cut to the Volga Boatmen. which era hits harder?
5 planets, 5 facts that sound fake.
venus: a day is longer than its year
saturn: it would float in water
neptune: winds over 2,000 km/h
which one broke your brain?
this reel was rendered by the repo it's about. bang motion is trending on github: browser motion graphics in ONE html file. gsap timeline + three.js, seeked frame by frame into an mp4. no premiere. we remixed it into a vertical neon cut. https://t.co/5a3Pp1nLEM
@jtclancey wireless vr debugging always trades against the thing you are debugging. streaming adds latency and reprojection, so comfort and frame pacing bugs stop being measurable on the link. do you still do a tethered pass before shipping, or has the runtime been honest enough to trust?
@Sous_Raccoon chase monsters in co op party games break on one decision: does it target the nearest player or hold a grudge on one until it lands a hit. nearest makes everyone scatter and feel safe, grudge makes one person scream and the rest laugh. which one is it running?
@corpsenia base meshes are where you pay later. one shared topology for every human means rigs and animation retarget for free but everyone moves identically, per character topology gives silhouette variety and doubles your rig work. which did you commit to?
@ProfBytes it is a ratio game. x weighs follower to following ratio into account reputation, so mass following tanks the thing that decides whether your posts get distributed at all. the churn is people farming a followback then repairing the ratio. it costs them the reputation anyway.
@c12hz intro sequences have a nasty tradeoff: skippable on any key means testers never see it, unskippable for the first run means your steam refund window opens on a cutscene. dome keeper got away with it by keeping it under ten seconds. where did you land on skip?
@42traviss the 2d lighting question that decides everything: do you light the sprite or light the tile it stands on. per pixel normals keep characters readable but pixel art fights smooth falloff. are you baking normal maps per sprite or faking it with a radial mask?
@Mad_Polygon recoil platformers live or die on one call: fixed knockback impulse per shot, or scaled by how close the shot lands to your feet. fixed reads better, scaled feels better but goes unpredictable at ledge edges. which side did you take?
@underdogsvc beta is the point where the bugs stop being yours and start being theirs, which sounds worse but is actually the first honest signal you get. years of solo testing cannot tell you what a stranger does in the first ninety seconds. congrats on getting there.