Public release - macOS 14+, Apple Silicon and Intel. Current build: 260621.
What stands out in v0.94
v0.94 builds on everything that happened from v0.90 through v0.93,
but the release story is different this time. The main headline is not a
single new surface feature. It is that the native build engine
is now effectively complete across the three target families that matter:
PiStorm, Classic Amiga, and
emulator workflows.
- Native build engine effectively complete - standard PiStorm, Classic, UAE, Amiberry, and MiSTer workflows can now run through the in-process Swift engine instead of leaning on the older script-heavy path.
- AmigaDiskKit now carries much more of the product - the same foundation now covers disk layout, filesystems, PiStorm FAT32 boot handling, file access, and more of the real build path.
- Far less glue in the middle - fewer handoffs between shell scripts, helper tools, and native code means cleaner verification and a much clearer long-term architecture.
- Greaseweazle stays the trusted real-floppy path today - while experimental DrawBridge support is now in, but still needs real-hardware validation before physical reads should be treated as reliable.
The native engine is now the release story
v0.90 introduced the first public AmigaDiskKit engine path. v0.92 made
Amiga Imager feel like a broader native toolkit. v0.93 pushed part of
that workflow into Finder. v0.94 is where the build pipeline itself
starts to feel fully native across the main supported targets.
- PiStorm, Classic, and emulator targets now share a much more unified in-process build architecture.
- The native path now covers more of the real work users actually rely on day to day, not just isolated formatting experiments.
- The result is a build workflow that feels more like one product and less like a native UI sitting on top of old scripting seams.
What this means in practice
This is not the flashiest release in visual terms, but it is one of the
most important engineering releases the project has had so far.
- Better architectural consistency - more of the disk, filesystem, and package flow now lives in one coherent native path.
- Cleaner future expansion - new platform work and new media workflows no longer need to begin from a shell-script-first baseline.
- Fallback where it still makes sense - a few excluded edge cases still keep script fallback, but the mainstream release story is now the native engine.
Floppy workflows: what ships now and what still needs validating
Greaseweazle remains the shipping native macOS floppy path
in Amiga Imager today. That is still the real supported route for reading,
writing, and working with physical Amiga floppies.
- Greaseweazle stays first-class - the native Swift floppy stack is still the current path for real drive work on macOS.
- DrawBridge is now in experimentally - it sits alongside Greaseweazle and expands the floppy side of the app, but it is not yet the path to rely on for physical reads.
- What still needs confirmation - the next real-hardware validation step is confirming the DrawBridge command bytes and 2-bit cell mapping before treating physical reads as trustworthy.
Why this matters
v0.94 is the point where the native build engine stops being only a
long-running internal migration and starts reading like the actual release
headline. It gives Amiga Imager a much stronger foundation for whatever
comes next, including broader floppy hardware support and more native media
workflows on macOS.