OpenCiv3: a Godot and C# remake of Civilization III built by modders

A remake aimed at the modding community, not at nostalgia

OpenCiv3, formerly known by the codename C7, is an open source remake of Civilization III built by the fan community with Godot and C#. The README describes it as mod-oriented, with capabilities inspired by other games in the 4X genre and lessons learned in Civ3 modding, and states the aim plainly: to create a new game that looks and plays like Civ3 while incorporating features from the best of the genre and the contributors' own ideas.

That last clause is the interesting one. The project is not a preservation effort. A preservation effort would try to reproduce the original exactly, bugs included, and would treat the original files as fixed input. OpenCiv3 instead wants to keep the feel while changing the rules where the modding community thinks they should change, which is a different and harder engineering goal than cloning a game.

The audience follows from that. It is for players and modders who liked the third Civilization and want a modern engine underneath, and for developers who want to study a large, real, open source project that deals with a genuinely awkward problem.

The project is hosted by the C7-Game organisation, with a homepage at openciv3.org, a subforum on CivFanatics, and a Discord server. The default branch is Development, and the last push was on 2026-10-20, with the repository not archived. The README notes the project is under active development and currently in a pre-alpha state, describing it as a rudimentary playable game that lacks many mechanics and late-game content, with errors likely.

That status line should be the first thing any prospective player reads, because it changes what the software is.

Reading Civ3 files is the hard part, and it is solved separately

The repository layout is where the engineering story is. Six directories carry the work, and four of them exist because of the 1990s data formats rather than because of the game rules.

QueryCiv3 is a data reader for Civ3 BIQ and SAV files that fetches data based on offsets from labelled section headers. Blast is an Apache-2.0 library for decompressing PKWare DCL, the compression used inside those BIQ and SAV files, and it was copied from an existing project rather than written from scratch. ConvertCiv3Media is a library for reading images and animations from Civ3, currently used by C7.

That separation is the key architectural decision. The legacy file formats are handled by dedicated, testable libraries with no dependency on the game, and the game itself consumes their output. Blast even has its own licence and a NOTICE file at the repository root, which is what you would expect from a project that vendors third party code.

The format work is unavoidable for a mod-oriented remake. A community that wants to keep the scenario editor, the saved games and the existing mod library working needs to read the original files, and the file formats are undocumented, compressed and offset-based. Doing that in one pass while also writing a game would produce an unreadable codebase.

The remaining directories follow the same logic. C7 is the core game running on the Godot engine, C7Engine holds the mechanics including the AI logic, EngineTests holds tests for the engine logic, and C7GameData stores native game data that will be saved to disk once the save feature is merged. The _Console directory and the doc directory sit alongside them.

One small note on the layout: the top-level file list does not include C7GameData, even though the README describes it, which suggests that directory is being reorganised on the Development branch.

Godot with C#, and a scripting language alongside it

The stack is worth naming precisely, because Godot is normally associated with GDScript and the choice here shapes how the project is written.

The README states that OpenCiv3 is built on the open source Godot Engine and written in C# and Lua. That pairing is the interesting part. Godot supports C# through its .NET build, and the project's primary language is listed as C#, so the game logic and the AI are typed, compiled and testable in the usual way, and the C7Engine and EngineTests split only makes sense if the engine code is ordinary C# that a test project can reference directly.

Lua then appears alongside it. The README does not say where Lua is used, and it would be a mistake to assume it drives the game rules. The most likely role, given that C# is the primary language and Lua is a scripting language embedded in the engine, is content and scenario scripting, but the documentation available here does not settle that, and the distinction matters if you want to extend the game. A modder who writes Lua expecting to script game logic, and finds that logic lives in compiled C#, has misunderstood the extension surface.

The default branch being Development rather than main or master is another signal. Combined with the pre-alpha status, it says the project expects breakage between releases, and that a user who wants a stable experience should take a tagged release rather than follow the branch.

The C7 directory is described as the core game that runs on the Godot engine, and C7Engine as the mechanics including AI logic. That split, with the engine isolated from the presentation layer, is what allows EngineTests to exist as a separate test project.

Installing means taking a tagged release, not cloning a branch

There are no build commands in the repository README, and that is a deliberate choice rather than an omission. The README says the latest stable version can be downloaded from the Releases page, and that current information on installation and features can always be found on the project homepage.

So the supported path for a player is a downloadable build, and the documentation of how to run it lives on openciv3.org rather than in the repository. A developer who wants to run from source needs the Godot .NET build for their platform, the C# toolchain, and the ability to open the project in the Godot editor.

The release history shows the shape of the versioning, and it is worth reading. There is a v0.3-dutch-preview-1 tag from 2025-12-07, a v0.2-carthage tag from 2025-01-05, and a v0.2-carthage-preview-4 tag with the subtitle Vanquishing the Zombies from 2022-11-05.

Two observations. The version numbers are pre-1.0, and the names are thematic, with Carthage and Dutch as the two named arcs, which fits a project that ships preview builds under narrative names. And the newest release, from December 2025, is a preview rather than a stable point release, which lines up with the pre-alpha description in the README.

The gap between the newest tag and the October 2026 last push is normal for this kind of project: work continues on the Development branch between named previews, and a player who wants the most recent code is effectively a developer rather than a user.

What pre-alpha actually means for a player

The README is unusually direct about the state of the software, and it is worth taking the sentence apart rather than reading it as a formality. It says OpenCiv3 is under active development and currently in a pre-alpha state, and describes it as a rudimentary playable game that lacks many mechanics and late-game content, with errors likely.

Three separate problems there. Lacking mechanics means rules are missing, not merely rough, and a strategy game with missing rules can be unplayable in a specific way, because the strategies that make the original interesting are often the rules themselves. Late-game content missing means the arc of a match is truncated, so a session ends before the interesting decisions appear. And errors are likely means the ones you hit may not be the ones anyone has reported.

The project handles this by giving the player a comparison point. The stated aim is to create a new game that looks and plays like Civ3, which is a testable claim, and the community links point at a CivFanatics subforum and a Discord where progress is discussed. The most useful thing a player can do with a pre-alpha remake is not to judge it as a game but to check it against the original for the specific mechanic in question.

The same caution applies to the modding story. The project positions itself as mod-oriented, and the presence of QueryCiv3 and ConvertCiv3Media says the ambition is real, but a mod built on a pre-alpha engine inherits whatever bugs exist at the time, and there is no compatibility promise in the README to fall back on.

MIT licensing, with a trademark line that matters

The code is under the MIT License, with a LICENSE file at the root and a NOTICE.md alongside it. The README states the project is a public open source project under the MIT licence, meaning anyone is allowed to contribute or make their own modified versions, and it links a Contributing page on the project wiki for the policies and guidance.

MIT is the right licence for this situation. It permits a company to base a product on the code, and it does not require modified versions to be shared back, though it does require keeping the copyright and licence notice, which for a game engine remake is the difference between a community project and a corporate blocker.

The NOTICE.md and the separate Blast directory licence are the two details that need care in practice. Blast is an Apache-2.0 library copied from another project, and Apache-2.0 carries its own notice requirements. If you redistribute a build that includes Blast, that licence and its notice travel with it.

The trademark position is the other half. The README states that OpenCiv3 is not affiliated with or endorsed by civfanatics.com, Firaxis Games, BreakAway Games, Hasbro Interactive, Infogrames Interactive, Atari Interactive or Take-Two Interactive Software, and that CIV and CIVILIZATION are registered trademarks of Take-Two Interactive Software, with all trademarks the property of their respective owners.

That disclaimer is standard, but it reflects a real constraint. The MIT licence covers the code the community wrote, not the rights to the Civilization name, and the original game's data and art are not the project's to license. A derived project should be careful with naming and with any redistributed original assets.

Why the C# and Godot combination is worth a look

Setting the game aside, the repository is a substantial open source codebase that solves a problem that is harder than it sounds, and that is a legitimate reason for an engineer to spend time in it.

Reading an undocumented binary format from a decades-old game, one that is compressed with PKWare DCL and organised around labelled section headers and offsets, is not a weekend exercise. QueryCiv3 and Blast do that work in isolation, and ConvertCiv3Media does the same for the graphics and animation data. The layering is the lesson: put the legacy format handling in libraries with no dependency on the game, then let the game consume typed structures.

The other thing worth studying is the split between presentation and rules. C7Engine holds the mechanics and the AI, C7 is the Godot-facing core, and EngineTests tests the engine logic. For a project with AI in it, being able to test the rules without starting the engine is worth arranging deliberately.

Compare this with the usual alternative for a remake, which is a single Unity or Godot project where file parsing, game logic and rendering live in the same assembly. The OpenCiv3 layout is more work up front and much better when a parsing bug turns out to be a rules bug.

The community context is also part of the appeal. The project is built by the fan community, its own channels are the CivFanatics subforum and a Discord, and the default branch is Development, which means the work is visible as it happens rather than arriving in annual releases.

Editorial conclusion

OpenCiv3 fits a Civ3 fan who wants a modern, moddable engine for the game, or a developer who wants to read a large C# and Godot project that solves the awkward problem of reading 1990s binary game files. It does not fit someone looking for a finished remake, and the project itself says mechanics and late-game content are missing and errors are likely. Before investing time, pull the latest release from the Releases page, run the same scenario under the original game and under OpenCiv3, and check the project's own status notes to see which mechanics are actually implemented.

Frequently asked questions

Official sources

1. C7-Game/OpenCiv3 on GitHub
2. License: MIT
3. Project website
4. README
5. Releases