Kamruz Moslemi’s 2011 Game Developer essay makes an argument that has aged remarkably well: the JRPG has been frozen since the 1980s, and the genre’s escape route was already built, shipped, and then ignored. His diagnosis is a coupled defect — random encounters exist only because the grind needs them, and the grind exists only because designers want a longer playtime number. Remove one and the other loses its justification. ...
Irmak “Max” Çavun, a 3D generalist working in AAA art production, set herself a constraint for a personal project: assemble a fully modular abandoned Middle Eastern courtyard in Unreal using only tiling textures and in-engine material blending, with no concept art. The reason was practical. Years of gathering reference for the region had collapsed into “shades of brown,” and she had been fooled more than once by AI-generated images sitting in her reference boards. ...
Nore Pollentier, a 19-year-old environment art student at Howest DAE, spent six weeks in a Sierra Division Academy mentorship recreating a small street in Ankara in Unreal Engine. It is a portfolio piece, so the framing is promotional — but the order of operations is a clean template, and the reasoning is stated rather than assumed. The project starts from target studios, not from an idea: Picked IO Interactive and MachineGames for their realistic, real-world-based environments; when IO had little public material, the pipeline research came from MachineGames artist talks. What those talks surfaced as priorities: material creation, modularity, sculpting, and matching the real location. Chose an existing street found on Pinterest over an invented concept, so composition was already solved and effort could go into realism. Reverse image search led to the exact spot on Google Maps for proportions from multiple angles. Reference grew throughout: PureRef boards, plus a habit of putting a scene screenshot next to the reference and playing “find the differences” — a cheap self-review loop worth stealing. Deliberate deviation budget: “even for a realistic piece, I still choose not to copy everything 1-to-1.” Production decisions that mattered: ...
80.lv interviewed CD PROJEKT RED’s Jakub Knapik (VP, Global Art Director) and Kacper Niepokólczycki (Environment Art Director) about how Night City was built. The headline answer is blunt: everything is hand-placed — every building, advert, trash container and sheet of paper on the sidewalk. The reasoning behind that is more interesting than the claim. Procedural tools were tested and rejected. “They didn’t even come close to the quality and complexity that we’re aiming for.” The one partial exception is a road tool that runs a spline through the world and auto-places roads, bridges and overpasses — and artists still adjust the results by hand. Four visual eras, one city. A week-long art directors’ offsite established Night City’s chronology: bleached 80s entropy, 60s-inspired plasticky kitsch, corporate neomilitarism, and rare pseudo-natural neo-kitsch for the ultra-rich. Districts differ by theme and colour coding, but they are built from those shared styles. The “rule of contrast.” A named team rule: put contrast into architecture, storytelling or vistas almost everywhere. Their example is a normal Rancho Coronado house with an AV crashed through the roof and NCPD investigating — absurd to the player, ordinary to the residents, which is what grounds the world. Density versus noise, admitted plainly. “In some places we did great, in some other places, less so.” The tiny-detail layer — adverts, neon, trash, facade colour — sometimes tipped into noise and hurt readability. Phantom Liberty answered with written guidelines for grouping colours, shapes and details and preserving negative space. First person changed the metrics. Cutscenes became interactive 360-degree environments, level-design metrics were rebuilt from scratch, texel density was doubled versus The Witcher 3 (later bought back with heavy decal use), and scale had to be established through familiar objects — paper, bottles, benches — since there is no visible character in frame. Lighting was designed as a sandbox. Probes and GI were pushed toward physical predictability and verified against offline Arnold renders before ray tracing existed, so RT and later path tracing arrived as an evolution rather than a rewrite. Under non-RT they budgeted roughly 10 active shadow-casting lights; path tracing put shadows on every loaded light to keep energy balance, forcing much more care about light placement. The postmortem lesson was organisational. The base game’s weak point was communication between discipline-based teams. The fix was flattening into multidisciplinary hubs around quests and the open world, keeping department leads as the backbone so feedback landed in one room and juniors could make direction-consistent calls. Both directors credit that for Phantom Liberty being among the smoothest projects of their careers. Two transferable takeaways: contrast rules and shared style vocabularies are what let hundreds of artists keep a decade-long world coherent, and readability at extreme density is an authoring problem — “controlled chaos” is composed, not generated. ...
Retro-gaming circles have argued for decades about which General MIDI module plays DOS game music “correctly.” John Novak settled it with recordings instead of opinions: 46 classic DOS soundtracks, rendered on 7 different MIDI modules — 322 lossless recordings, all published on the Internet Archive along with the MIDI files and REAPER projects. The method is the point. The same captured MIDI stream was fed to every module, so the synthesizer is the only variable: ...
Bruno Croci wrote a Voronoi-noise fragment shader as a background for a music video. It ran fine on an Intel iGPU laptop, a Pixel 9 Pro and an RTX 2070 — and stuttered on exactly one machine: a Windows PC with an RTX 4070. The investigation that followed is a clean model of how to debug “the compiler did something weird.” The one-line clue came before any deep dive: changing the scalar multiplying the noise input from 398.0 to 398.1 made the stutter disappear. ...
Caroline Esmurdoc joined Double Fine in 2003 as executive producer, hired to take over a project that was already two and a half years old and had never had a producer. Her postmortem ran in the August 2005 issue of Game Developer magazine, and it reads less like a design retrospective than a business one. The project numbers set the tone: $11.8 million budget, 42 full-time developers at peak 166,781 lines of C++ for the game, 332,650 lines of Lua across 2,433 files 4.5 years of calendar development — “actually developed in less than two” The team’s own summary of the result: they dodged a hundred bullets without compromising the game, losing ownership of the company, or missing a day of payroll. ...
Tommy Thompson’s breakdown of Alien: Isolation draws on a 2016 nucl.ai talk and follow-up questions with Andy Bray at Creative Assembly. The problem it answers is specific: how do you get one unkillable antagonist to carry a 12–18 hour horror campaign without scripting the scares? Coordinated jump scares work once. The design goal was Bray’s phrase “psychopathic serendipity” — the alien always seems to be in the right place at the right time, without ever being scripted to be. ...
Factorio’s quality mechanic rolls for higher item tiers, and Wube’s framing is that it all comes down to statistics: craft enough and the distribution does the work. Gegell, a CS grad student working in 3D computer vision, spent two years proving there is another route — read the game’s random number generator out of crafting observations, predict thousands of rolls ahead, and skip the RNG forward until a legendary item is guaranteed. ...
Joshua Barretto, one of Veloren’s core developers, wrote up the choices his team made that run against the grain — the open-source multiplayer voxel RPG started in 2018 and is still going. Most of the post is about what those decisions cost, which is what makes it worth reading. ECS from day one. Started before ECS was common outside demoware, so the team invented internal concepts to make it work. Payoff: 50% core utilisation on a 48-thread server with 500+ players and tens of thousands of interacting entities — without sharding or shrinking gameplay scope. ECS inverts the default. In a class hierarchy, polymorphism is opt-in; with an ECS it is the default and taxonomy is what you opt into. Consequences included players getting an ItemDrop component for their carried items and being able to pick each other up, and mount cycles producing towers of entities riding each other. Players and NPCs obey identical rules. NPCs use the same physics, the same Controller component (a virtual gamepad fed by keyboard for players and a decision tree for NPCs), the same movement code, and the same skill tree. If the agent code ignores momentum at a cliff edge, the NPC falls off. Voxels are stored as “chonks”. Not big arrays, not RLE, not octrees — an index table marking NxNxN groups homogeneous or heterogeneous, split into vertical sub-chunks. Trades compression against cache coherence, keeps random access fast. The world is pre-generated. Full map at low resolution on startup, details filled in near the player. That is what makes long rivers always flow downhill possible; local constraint solving cannot express it. It also gives free level-of-detail for distant terrain. Generation is ontological, not teleological. Rather than judging outputs by whether they look right, Veloren models physical inputs and simulates them — hydraulic erosion for terrain and rivers, a traversal-cost model for paths. RTSim never stops. Leaving a player’s view does not despawn an NPC; it gets a low-resolution dual that keeps running its decision tree. Quests, faction dynamics and parts of the economy live there, tracking tens of thousands of NPCs at once. No invisible walls, one world space. Pulling a boss out of its dungeon into a town is allowed, which forces systems to be designed defensively. No loading screens means cave networks a kilometre deep must not leak surface lighting — global effects like lightning keep finding ways through shadow maps and screen-space reflections. The best paragraph in the post is the aside on what procedural generation actually is: not random content, but constraints that produce self-consistency. Good generators need almost no randomness, because randomness is what the player brings — the generator’s job is to coerce it into a system with consequences. If you find a river you should be able to follow it to the source; if you kill a monster, the nearest town should notice. ...
A dev in a PhD program posts a game they are proud of to a friendly sub and a general solo-dev sub. The first is generous; the second turns in a day, and the comments become mostly “AI slop.” They could not have made the game any other way, and the ask is not “how do I win the argument” — it is how anyone copes with this. The replies that are worth reading skip the encouragement and try to name the mechanism. The consensus diagnosis is not really about art quality: ...
James Nouch opens by admitting he reloads the moment a persuasion check fails, then asks why digital RPG failure reads as punishment while failure at a tabletop is often the best part of a session. He put the question to four designers: Colin McComb (Planescape: Torment, Fallout 2), George Ziets (Wasteland 3, New Vegas), Gareth Damian Martin (Citizen Sleeper) and Christoffer Bodegård (Esoteric Ebb). The diagnosis: McComb: players reload because they were taught that a failed roll means they screwed up — and the systems agree, taking gold, items, XP and checkpoint time Ziets: Planescape: Torment’s mortuary was the counter-example. Die, wake up, and characters react; you see narrative you would never have seen had you reloaded Damian Martin: it is not that players want to conquer everything, it is that games insist victory is always possible and the player’s power must be respected — the “supercop-slash-God” problem Bodegård: the standard power fantasy builds something that ends up hollow — “a bit masturbatory” What they build instead: ...
ByerN writes games after hours — 2–5 hours on an average work day, alongside a Java backend job — in an unusual stack of libGDX and Scala. Over three years he shipped three node-based games solo and self-published on Steam, and then published the actual numbers rather than just the lessons. Node Farm — ~12 months, released Nov 2024, 16,504 units, $102,222 net lifetime Node Math — ~8 months, released Oct 2025, 12,661 units, $60,535 net Node Crown — ~8-9 months, finished, unreleased, 676 wishlists The plan behind them is a franchise bet: one shared codebase, one new experimental mechanic per game, features backported across games. Node Math took eight months partly because nothing had to be reinvented. ...
Tim McGraw, an associate professor at Purdue, and his student Jack Myers built a system that turns cryosection imagery — including Visible Human Project data — into virtual bodies you can cut, tear, retract, and separate in real time, on ordinary GPUs. It won the Audience Choice award at SIGGRAPH 2026’s Real-Time Live!. The interview is a walkthrough of one representation decision and everything that follows from it. Existing anatomy tools already use this data, but only to clip it against a plane or make structures transparent. McGraw’s problem is interactivity: deforming and cutting the way you would in a dissection lab. ...
Pavel Kuzmenko, a technical artist and tools developer, wrote texLab: a native C++ plugin that packs PBR maps into ORM/RMA-style layouts inside Unreal Engine 5.7 and 5.8, and unpacks packaged textures back into individual maps. The pitch is deleting a round trip — export maps, open another DCC, pick a channel layout, pack, reimport, recheck — where the texture on disk may no longer match the preview you approved. The interesting part isn’t the packing. It’s the decisions that decide whether artists trust the tool. ...
Daniel “Agentlien” Kvick, a graphics programmer at 505 Games, spent last year writing texture-conversion code to port a custom PC engine to consoles. He went in expecting a chore and came out with a tour of how much platform-specific machinery is packed into the word “texture.” His framing is a debugging problem: given a block of memory holding a texture, converting it means computing, for every element in the hierarchy, the address it will occupy after conversion. Each layer has its own ordering rule, and the rules differ per platform. Running example: a 1024x1024 RGBA texture in BC7. ...
nand2mario spent a month adding features to z486, his 486-class FPGA CPU core, and then went looking for the moment PC gaming changed — the arrival of 3D accelerators. The card that did it for him was the 3dfx Voodoo, seen through Need for Speed II SE. Could it be recreated in programmable logic now that the CPU existed? The result is zSST, a SystemVerilog implementation of Voodoo Graphics (SST-1), combined with z486 into z486 XL: a DOS PC with Voodoo graphics running inside a Xilinx KV260. Tomb Raider runs with its original Glide renderer. ...
“Deprofessionalization” is the clumsy industry word for a real shift: fewer career roles at big studios while small and solo teams ship games that compete with them. Bryant Francis, who helped popularize the term, argues the trend has a second edge. After reporting on sexual harassment allegations against three indie horror publishers, he started noticing the pattern as a workplace problem rather than an economic one. His point is not that indies are uniquely bad. It is that casual environments remove the structures that name bad conduct: ...
Disney spent decades promising “Living Characters” — animatronic or digitally puppeteered figures that could interact with park guests the way NPCs do in a game. Defunctland’s four-hour video essay treats the initiative as a broken promise, and Bryant Francis pulled three design lessons out of the wreckage for Game Developer. The most useful one cuts against the instinct to chase fully autonomous agents. The examples come from what actually shipped: ...
Entity-Component-System architecture grew up in game engines; relational databases grew up in enterprise software. They never compared notes. Loïc Baumann — who worked on Unity’s DOTS and is now VP of engineering at MEGA — lays out how the two converged anyway, and argues that game servers are the place the mismatch finally has to be resolved: they need engine-grade throughput and database-grade guarantees in the same process. His mapping is the whole thesis in five lines: ...