NitrOS 9's History

NitrOS-9's History

As the decades pass, pieces of NitrOS-9 history can easily be lost. This page exists to preserve some of that story, even if it is incomplete and assembled from multiple voices.

L. Curtis Boyle

According to Curtis, NitrOS-9 began with:

• Bill Nobel
• Wes Gale
• L. Curtis Boyle

Alan DeKok joined later.

Alan's "OS-9 Level 3" work focused on expanding system memory. Curtis notes that it did not work with DriveWire in its current form, though Bill Nobel had looked into that at one point.

He also points back to the late-1980s "Version 3" effort, a much larger project involving more than a dozen developers and intended for Tandy distribution. Kevin Darling led that effort, and Brother Jeremy later released one version of it. Parts of that work eventually influenced NitrOS-9:

• some pieces were ported directly, such as GShell and serial mouse drivers
• some features were recreated in clean-room form
• some ideas never made it into the current tree

Curtis also points to a series of group interviews that capture the project's history across different eras:

• Part 1 (aired August 20, 2022): Episode 275 - First decade of NitrOS9
• Part 2 (aired November 2, 2022): Episode 286 - Second decade of NitrOS9
• Part 3 (aired May 27, 2023): CoCo Nation Show #314

Those interviews include contributors from the original and open-source eras and are a strong starting point for anyone wanting a broader oral history.

Boisy Pitre

Boisy traces his involvement back to around 1998, when he got Alan DeKok's TuneUp, a collection of patches intended to make OS-9 Level Two run faster on the CoCo 3.

At that point there were already many scattered patches, enhanced commands, and module improvements online. That led to a larger idea: build a complete OS-9 system for the CoCo 3 directly from source.

That effort started with disassembly work. Early source was only lightly commented, but it provided a foundation that could be improved over time.

Later, Boisy incorporated the enhanced 6309 modules from NitrOS-9 and added conditional assembly so both 6309 and 6809 builds could be supported from the same source base.

Cross-Development Tools

This source-based approach naturally led to cross-development tools, including:

• mamou, a 6809 cross-assembler
• Toolshed, for manipulating disk images

Those tools made it possible to build the entire operating system on a modern computer and then produce bootable media for real hardware.

At the time, Boisy was developing primarily on Linux, so those tools started there.

Source Control and Collaboration

Around 2000, the work began moving under source control, and more collaborators joined the effort to improve disassemblies and comments.

The work also expanded beyond CoCo 3 Level Two:

• OS-9 Level One modules for the CoCo 1 and CoCo 2 were disassembled
• consolidation between Level One and Level Two source trees began

That marked the start of a more unified, maintainable source base.

DriveWire Connection

Boisy also notes that this work helped bootstrap DriveWire.

Once full system builds and disk-image tooling existed, it became natural to move those disk images over a serial connection to a modern machine. That idea led directly into the early DriveWire work around 2003, during his time in Boston.

By then, he had moved to the Mac, which became his main development platform.

Why This Page Matters

NitrOS-9 did not appear all at once. It grew out of:

• earlier OS-9 enhancement efforts
• disassembly and reconstruction work
• cross-development tooling
• collaboration across multiple eras of contributors

This page captures a few of those threads and gives future contributors a starting point for understanding how the project evolved.