TL;DR
Get the latest gadgets delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
A CedarDB developer has ported the original 1993 Doom game logic and renderer to SQL and says the game runs inside a database. The implementation runs the game loop at 35 frames per second, can render frames at up to 60 Hz on the developer’s laptop, and supports online deathmatch.
A CedarDB developer has ported the game logic and renderer of the 1993 game Doom to SQL, running the game inside a database rather than relying on a conventional game engine. The developer says the implementation runs Doom’s game loop at its original 35 frames per second, renders frames at up to 60 Hz on a laptop, and supports multiplayer deathmatch.
The project, called SQLDoom, keeps game state, gameplay logic and rendering in the database. A Python client handles keyboard input, advances the game at the required rate and displays the resulting image. According to the developer, it does not calculate gameplay or draw the frame itself: SQL queries perform those tasks, returning a table or bitmap containing the image’s RGB pixel values.
The developer describes the game loop and renderer as separate processes. The game logic advances at a fixed 35 Hz, while the client can request rendered frames independently. The result, the developer says, preserves recognizable Doom gameplay, including movement, combat, enemy behavior, items, projectiles, explosions, animation, view bobbing and the heads-up display.
The report also says four-player deathmatch is available on European and US servers, using the shareware version’s first episode. Players can join a queue when all four slots are occupied; if that queue is full, the project offers access to live game state through SQL queries. These availability details and performance figures come from the developer’s project report, not an independent benchmark.
What SQLDoom Shows About Databases
SQLDoom is a technical demonstration of how far database queries can be pushed beyond storing and retrieving records. Doom requires continuous updates to player and enemy states, collision and combat calculations, and the production of a complete image. Implementing those tasks in SQL makes the project an unusual test of database capabilities, rather than simply a way to display a map or game-related data.
The work may interest database developers and game programmers as an experiment in query-based computation. It does not establish that SQL is a practical replacement for conventional game engines: the report does not provide comparative performance tests, operating costs or evidence that the approach scales to larger or more demanding games. Its immediate significance is that the developer says a recognizable, playable Doom implementation—including multiplayer—can run with the database doing the game logic and rendering.
As an affiliate, we earn on qualifying purchases.
From ASCII Demo to Full Doom
The project follows an earlier experiment by the same developer, DOOMQL, which produced a rough ASCII-style view of Doom at about 30 frames per second. The developer says that version used raycasting, an approach more closely associated with Wolfenstein 3D, rather than Doom’s rendering method. The new work was intended to reproduce the original game’s rendering and play, not just create a scene that resembles it.
The report identifies binary space partitioning (BSP) as part of what distinguishes Doom’s rendering approach. It also describes converting Doom’s WAD game-data files into relational database tables. The developer says importing all of Doom 1’s data takes about 18 seconds on a laptop and required roughly 1,000 lines of Python. That Python is used for importing data as well as the client functions; according to the project’s stated rules, gameplay and rendering remain in SQL.
“The original Doom’s game logic and renderer, both implemented as SQL queries.”
— SQLDoom developer, in the CedarDB project report
Performance and Availability Questions
The published figures are the developer’s own account. The report does not give a detailed hardware-independent benchmark, disclose performance across different database configurations or compare SQLDoom’s resource use with a conventional Doom port. The claim of rendering at up to 60 Hz is specifically tied to the developer’s laptop, while the game loop is reported at 35 Hz.
The report describes deathmatch servers in Europe and the United States, but does not specify how long they will remain available, how many players are using them, or what network latency and server capacity they support. It also does not establish that every part of the original Doom release or its broader multiplayer ecosystem has been reproduced. The stated playable version is the first episode of the shareware game.
Testing the Live SQLDoom Build
The developer’s report directs readers to live SQLDoom servers, where four-player deathmatch is offered and queued players may inspect game state through SQL. The project’s immediate next test is whether readers can access those servers and reproduce the reported experience under their own conditions.
No broader release schedule, independent evaluation or follow-up benchmark is specified in the report. Further details about performance, database requirements and server availability would help establish how portable the demonstration is beyond the developer’s setup. Until then, SQLDoom is best understood as a reported technical port and playable experiment, not evidence that databases are ready to replace purpose-built game engines.
Key Questions
What is SQLDoom?
SQLDoom is a project that ports the original 1993 Doom’s game logic and renderer to SQL and runs them inside a database. The developer says a Python client handles input, timing and displaying the rendered image.
Does the game really run entirely in SQL?
According to the developer, the game logic and rendering run as SQL queries. The implementation still uses Python to import game data and to operate the client, which reads keyboard input, advances the game loop and displays the output.
How fast does SQLDoom run?
The developer reports a 35 Hz game loop, matching the original Doom’s stated rate, and rendering at up to 60 Hz on a laptop. These are project-reported figures, not results from an independent benchmark.
Can people play SQLDoom online?
The report says four-player deathmatch is available on servers in Europe and the United States. It identifies the playable content as the shareware version’s first episode and says players can queue when all game slots are occupied.
Does SQLDoom show that SQL can replace game engines?
No. It shows that the developer has implemented a demanding game as a database-based experiment. The report does not include comparisons or broader performance evidence that would show SQL is a practical replacement for conventional game engines.
Source: hn
Fall Picks
fall essentials
As an affiliate, we earn on qualifying purchases.
