Skip to main content

VernonDSL

VernonDSL is a tensor-first Python language and MLIR compiler for GPU graphics and compute. A restricted, statically typed Python frontend lowers through shared semantic IR into target-specific CPU, CUDA, Vulkan, OpenGL, DirectX, and Metal artifacts.

For an end-to-end explanation of the language, compiler backends, Runtime/RHI, ExecutionGraph, and offline cooking model, see ARCHITECTURE.md.

from typing import Annotated

import vernon_dsl as vd

WIDTH = 640
HEIGHT = 320


@vd.func
def complex_square(z: vd.Vector[vd.f32, 2]) -> vd.Vector[vd.f32, 2]:
    return vd.Vector([z[0] ** 2 - z[1] ** 2, z[1] * z[0] * 2])


@vd.kernel(workgroup_size=(16, 16, 1))
def paint(
    pixels: vd.TensorView[vd.f32, (vd.dyn, vd.dyn), vd.write],
    time: vd.f32,
    gid: Annotated[vd.Tensor[vd.u32, (3,)], vd.builtin("global_invocation_id")],
) -> None:
    x = gid[0]
    y = gid[1]
    if x < WIDTH and y < HEIGHT:
        c = vd.Vector([-0.8, vd.cos(time) * 0.2])
        z = vd.Vector(
            [
                (vd.f32(x) / vd.f32(HEIGHT) - 1.0) * 2.0,
                (vd.f32(y) / vd.f32(HEIGHT) - 0.5) * 2.0,
            ]
        )
        iterations = 0
        while vd.norm(z) < 20.0 and iterations < 50:
            z = complex_square(z) + c
            iterations += 1
        pixels[y, x] = 1.0 - vd.f32(iterations) * 0.02


vd.init(arch=vd.cuda)
pixels = vd.storage.zeros(dtype=vd.f32, shape=(HEIGHT, WIDTH))
paint(pixels, 0.0, grid=(WIDTH, HEIGHT, 1))
image = pixels.to_numpy()

The current stable release is 0.1.1. It supports CPython 3.11 through 3.14 on Windows x64, Linux x64, and Apple Silicon macOS. The frontend remains language version 3 while the v4 roadmap is developed. See the 0.1.1 release notes, PUBLIC_API.md, and COMPATIBILITY.md for the supported surface and compatibility contract.

Demos

VernonDSL Julia set
Julia Set
Tensor compute, structured control flow, and device readback
VernonDSL ray-marched terrain
Ray-marched Terrain
VernonDSL animated Mandelbulb
Animated Mandelbulb

The Julia Set showcases tensor compute. Terrain and Mandelbulb exercise real-time fragment pipelines, ray marching, texture sampling, mathematical intrinsics, offscreen rendering, and host readback.

From a source checkout, install the example dependency and run an interactive 60 FPS presentation:

$env:PYTHONPATH = "$PWD/python"
uv sync --extra examples --frozen

uv run --frozen --no-sync python examples/terrain_showcase.py `
  --arch vulkan --preset showoff --fps 60

uv run --frozen --no-sync python examples/mandelbulb_showcase.py `
  --arch vulkan --preset showoff --fps 60

Press Escape or Q to exit. Use --arch directx or --arch opengl on Windows when Vulkan is unavailable. The presenter targets the requested frame rate; actual throughput depends on the GPU, driver, image size, and readback cost. Use --preset smoke --headless for fast acceptance checks.

Install from PyPI

Prebuilt wheels are provided for Windows x64, Linux x64, and Apple Silicon macOS 15 or newer for CPython 3.11 through 3.14:

py -m pip install vernon-lang==0.1.1

VernonDSL 0.1.1 is wheel-only. Intel macOS, source distributions, PyPy, and other Python versions are not published.

Verify the installation:

py -c "from importlib.metadata import version; print(version('vernon-lang'))"
vernon-compile-python --help
vernon-cook-pipeline --help

The wheel contains the Vernon compiler, CPU Runtime, and the native backend implementations supported by its platform. It does not bundle GPU drivers. Runtime availability therefore depends on the selected backend:

  • CPU: compute reference execution with no additional system dependency;
  • CUDA: compute with a compatible NVIDIA driver;
  • Vulkan: compute and offscreen graphics with a loader and vendor ICD;
  • DirectX 12: compute and offscreen graphics using Windows and its GPU driver;
  • OpenGL: compute and graphics through a compatible system context;
  • OpenGL ES: compute and graphics through a compatible owned or external context;
  • Metal: compute and offscreen graphics through the system framework on supported Apple Silicon Macs, with no additional loader.

Cooked MSL bundles are consumed by the Runtime on Apple. Metal presentation and swapchain management are outside the 0.1.1 contract. Argument-buffer pipelines fail explicitly when the selected device cannot provide the required tier or encoder.

CPU graphics, CUDA images and samplers, f16/f64 vertex attributes, non-relaxed atomics, asynchronous dispatch, and multiple frames in flight are outside the supported 0.1.1 subset. See RELEASE_NOTES.md for the complete release contract.

Optional Vulkan setup

macOS

Metal is the zero-install GPU backend on macOS. Vulkan is optional because macOS does not provide it natively. To use Vulkan, install the Khronos loader and MoltenVK ICD with Homebrew:

brew install molten-vk vulkan-loader

Vernon also searches Homebrew locations under /opt/homebrew and /usr/local. VERNON_VULKAN_LOADER may name the installed Khronos loader dylib; do not point it directly at MoltenVK.

Linux

Linux users need only the driver stack for the backend they select. For Vulkan on Ubuntu or Debian, install the loader and an appropriate vendor ICD. Mesa provides Intel, AMD, and software Vulkan drivers:

sudo apt-get update
sudo apt-get install --yes libvulkan1 mesa-vulkan-drivers

NVIDIA systems should install the matching proprietary driver instead of relying on Mesa for the device ICD.

Loader discovery

Vulkan is discovered when the Runtime creates a device. Vernon tries VERNON_VULKAN_LOADER, a loader under VULKAN_SDK, and the platform loader name.

VERNON_ENABLE_VULKAN_RUNTIME controls whether Vulkan support is included in the build. It does not indicate that a loader, ICD, or usable device is present on the machine running Vernon. Device creation reports the attempted loader locations and Vulkan initialization error when runtime discovery fails.

Cook deployable artifacts

Declare a persistent asset beside its shader stages:

mesh_asset = vd.pipeline_asset(
    id="pipeline/mesh",
    program=(mesh_vertex, mesh_fragment),
    variants=((), (INSTANCE,), (SKIN,), (INSTANCE, SKIN)),
)

Cook it for a deployment target without importing or executing the source module:

vernon-cook-pipeline examples/variant_mesh.py:mesh_asset `
  --target vulkan `
  -o build/variant_mesh

For a CPU compute asset:

vernon-cook-pipeline python/tests/pipeline_asset_fixture.py:scale_asset `
  --target cpu `
  -o build/cpu_scale

The output contains a versioned *.pipeline.json manifest and content-addressed files under artifacts/. Depending on the target, artifacts are SPIR-V, GLSL/ESSL, DXIL, PTX, Metal source, LLVM IR, or relocatable CPU objects. Cooked Metal bundles contain MSL consumed by the Runtime on Apple. Missing variants and unsupported target combinations fail explicitly rather than silently falling back.

Pipeline sources, MLIR, manifests, shader artifacts, native objects, and caches are executable input, not sandboxed data. Do not compile or load untrusted artifacts without isolation; see SECURITY.md.

When working from a source checkout where tool.uv.package = false, invoke the cooker as a module:

$env:PYTHONPATH = "$PWD/python"
uv run --frozen --no-sync python -m vernon_dsl.pipeline_asset_cli `
  examples/variant_mesh.py:mesh_asset --target vulkan -o build/variant_mesh

Build from source

Source and wheel builds are continuously tested on Windows, Linux, and macOS. Required tools:

  • Git, CMake, and a platform C++ toolchain;
  • Visual Studio 2022 C++ tools and the Windows SDK on Windows;
  • Python 3.11 through 3.14 and uv;
  • the repository's pinned llvm-project submodule.

For a full Ubuntu or Debian source-build and graphics-test environment, install the same native packages used by Linux CI:

sudo apt-get update
sudo apt-get install --yes \
  ninja-build patchelf pkg-config \
  libvulkan1 mesa-vulkan-drivers \
  libgl1-mesa-dev libegl1-mesa-dev \
  libwayland-dev libxkbcommon-dev wayland-protocols \
  xorg-dev xvfb

Initialize the repository and build the pinned LLVM/MLIR installation once:

git submodule update --init --depth 1 llvm-project
python scripts/configure_llvm.py --build

The helper selects the LLVM target for the host architecture and reuses the generator recorded in an existing build directory. Otherwise it selects Visual Studio 2022 on Windows, Ninja when available, or Unix Makefiles. CUDA/NVPTX is enabled when libdevice.10.bc is found through CUDA_PATH, CUDA_HOME, CUDAToolkit_ROOT, CONDA_PREFIX, nvcc, or the vendored llvm-project/nvidia-nvcc directory. Use --cuda off to disable CUDA or --cuda on --libdevice PATH to require it. Omit --build to configure only; run python scripts/configure_llvm.py --help for all overrides.

Build VernonDSL:

On macOS:

mkdir osx_build
cd osx_build
cmake ..
cmake --build . --parallel
ctest --output-on-failure

On Linux:

mkdir linux_build
cd linux_build
cmake ..
cmake --build . --parallel
ctest --output-on-failure

On Windows PowerShell:

mkdir windows_build
cd windows_build
cmake ..
cmake --build . --config Release --parallel
ctest -C Release --output-on-failure

CMake runs the frozen uv sync, uses MLIR and LLD from llvm-project/install, selects the synchronized Python interpreter and host compiler architecture, and disables runtime backends unsupported by the target platform and architecture. Single-configuration generators default to Release. Tests and the staged-file formatting Git hook are enabled by default. Each setting remains available as a -D override.

CMake places the development _native and _gl_context modules in python/vernon_dsl/. Source-checkout Python commands therefore use PYTHONPATH=python and uv run --frozen --no-sync.

Build a release wheel:

uv build --python 3.11 --wheel --no-cache --clear
uvx --from twine twine check dist/*.whl

On macOS, set MACOSX_DEPLOYMENT_TARGET=15.0 for the supported arm64 wheel, matching the release CI.

Clean wheel builds fetch pinned third-party CMake dependencies. They can take several minutes on the first run.

Run tests

Run the Python suite:

$env:PYTHONPATH = "$PWD/python"
uv run --frozen --no-sync pytest python/tests -q

Run native tests:

ctest --test-dir osx_build -C Release --output-on-failure

Use linux_build or windows_build instead when following the corresponding platform build example above.

Run the release coverage gates:

$env:PYTHONPATH = "$PWD/python"
uv run --frozen --no-sync coverage erase
uv run --frozen --no-sync coverage run -m pytest python/tests -q
uv run --frozen --no-sync coverage json -o coverage.json
uv run --frozen --no-sync python scripts/check_python_coverage.py coverage.json

Run GPU-optional showcase acceptance:

$env:PYTHONPATH = "$PWD/python"
uv run --frozen --no-sync python scripts/run_showcase_smoke.py `
  --output-dir build/showcase-smoke `
  --summary build/showcase-smoke/summary.json

Unavailable GPU backends are reported as explicit skips. Use --required when every requested backend must execute.

Run examples

All source examples use the same development environment:

$env:PYTHONPATH = "$PWD/python"
uv sync --extra examples --frozen

Headless showcase output:

uv run --frozen --no-sync python examples/terrain_showcase.py `
  --arch vulkan --preset showoff --headless `
  --output build/terrain.png --result-json build/terrain.json

uv run --frozen --no-sync python examples/mandelbulb_showcase.py `
  --arch vulkan --preset showoff --headless `
  --output build/mandelbulb.png --result-json build/mandelbulb.json

Compute-rendered Julia set:

uv run --frozen --no-sync python examples/fractal.py --arch cuda
uv run --frozen --no-sync python examples/fractal.py --arch vulkan
uv run --frozen --no-sync python examples/fractal.py `
  --emit-metal build/fractal.metal

End-to-end compute and graphics pipeline:

uv run --frozen --no-sync python examples/complete_pipeline.py `
  --arch vulkan --frames 3 --headless `
  --output build/complete-color.png `
  --id-output build/complete-object-id.png

More examples and their third-party attributions are documented in examples/README.md.

Future roadmap

The post-0.1.1 roadmap includes:

  1. finish and accept the language-v4 contract, including first-order pure-function autodiff;
  2. broaden repeatable hardware-backed GPU acceptance;
  3. design the ABI change required for asynchronous dispatch and deferred multi-frame resource reclamation;
  4. replace the remaining temporary compiler bridges tracked in the completion roadmap.

Release support and security reporting are documented in SUPPORT.md and SECURITY.md. Published wheel, checksum, SBOM, and provenance requirements are documented in RELEASE_ARTIFACTS.md.

Related design and release documents:

License

VernonDSL is licensed under the Apache License 2.0. Third-party notices for adapted showcase shaders are retained under examples/.

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

vernon_lang-0.1.1-cp314-cp314-win_amd64.whl (45.3 MB view details)

Uploaded CPython 3.14Windows x86-64

vernon_lang-0.1.1-cp314-cp314-manylinux_2_35_x86_64.whl (55.9 MB view details)

Uploaded CPython 3.14manylinux: glibc 2.35+ x86-64

vernon_lang-0.1.1-cp314-cp314-macosx_15_0_arm64.whl (43.8 MB view details)

Uploaded CPython 3.14macOS 15.0+ ARM64

vernon_lang-0.1.1-cp313-cp313-win_amd64.whl (44.0 MB view details)

Uploaded CPython 3.13Windows x86-64

vernon_lang-0.1.1-cp313-cp313-manylinux_2_35_x86_64.whl (55.9 MB view details)

Uploaded CPython 3.13manylinux: glibc 2.35+ x86-64

vernon_lang-0.1.1-cp313-cp313-macosx_15_0_arm64.whl (43.8 MB view details)

Uploaded CPython 3.13macOS 15.0+ ARM64

vernon_lang-0.1.1-cp312-cp312-win_amd64.whl (44.0 MB view details)

Uploaded CPython 3.12Windows x86-64

vernon_lang-0.1.1-cp312-cp312-manylinux_2_35_x86_64.whl (55.9 MB view details)

Uploaded CPython 3.12manylinux: glibc 2.35+ x86-64

vernon_lang-0.1.1-cp312-cp312-macosx_15_0_arm64.whl (43.8 MB view details)

Uploaded CPython 3.12macOS 15.0+ ARM64

vernon_lang-0.1.1-cp311-cp311-win_amd64.whl (44.0 MB view details)

Uploaded CPython 3.11Windows x86-64

vernon_lang-0.1.1-cp311-cp311-manylinux_2_35_x86_64.whl (55.9 MB view details)

Uploaded CPython 3.11manylinux: glibc 2.35+ x86-64

vernon_lang-0.1.1-cp311-cp311-macosx_15_0_arm64.whl (43.8 MB view details)

Uploaded CPython 3.11macOS 15.0+ ARM64

File details

Details for the file vernon_lang-0.1.1-cp314-cp314-win_amd64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp314-cp314-win_amd64.whl
Algorithm Hash digest
SHA256 1c92a1e2491f844d000131bb6549d0f975dfdc4c6b6c56cac474ac40706c2977
MD5 68ea4748cd8ea58ec457793f8cca531f
BLAKE2b-256 27094bb098d80d953e73f93f063c5f72e1a8bf39a382818c69074ce234f7f54f

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp314-cp314-win_amd64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp314-cp314-manylinux_2_35_x86_64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp314-cp314-manylinux_2_35_x86_64.whl
Algorithm Hash digest
SHA256 dc35af7c0251fdb37aca35d492f9d915893e611a921bdb1013f8fffdaabfd1ac
MD5 61b3e8a632ce3eeb9cf8e791528684fd
BLAKE2b-256 3371615c67c9f6333627a69820b167a97270b4073186d804bffd337dfd28a423

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp314-cp314-manylinux_2_35_x86_64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp314-cp314-macosx_15_0_arm64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp314-cp314-macosx_15_0_arm64.whl
Algorithm Hash digest
SHA256 3bcece3310f61104bba948ef16816e7aabbeff7f1283db9c1a7ff794f0205c6e
MD5 64a5c1cefbdb6600fd60907659805297
BLAKE2b-256 b86862be0045f48892c9d88ac546be8527d168483c10dcd5d8464cf1ba689ddf

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp314-cp314-macosx_15_0_arm64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp313-cp313-win_amd64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp313-cp313-win_amd64.whl
Algorithm Hash digest
SHA256 71a302f409138f04c6d6090ccd4a41d3ef516ad3d1b701c98e14ec7da585f474
MD5 f483da70211eeb4782a5d2d09e0c995d
BLAKE2b-256 c6b9e40fe07ce7dcfa84c079b71f4c8b414d54c617de3c123432248a4365f1c3

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp313-cp313-win_amd64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp313-cp313-manylinux_2_35_x86_64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp313-cp313-manylinux_2_35_x86_64.whl
Algorithm Hash digest
SHA256 e3607d9d8b648197d777b5f28094bde5918e9f80553aa23983f451c1d3bbea44
MD5 f2db4f80790fea4ac39909a249c0c34c
BLAKE2b-256 5eca220e202dccdc2d1b9f6aff6054f998305f20b4a683a7364c18e7c1afb37f

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp313-cp313-manylinux_2_35_x86_64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp313-cp313-macosx_15_0_arm64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp313-cp313-macosx_15_0_arm64.whl
Algorithm Hash digest
SHA256 370fdb2f851676770cd0e668167f0988e8e2aff801fa25361995ff0ac079887e
MD5 12eed912fc80349b4b47a702ce626bdb
BLAKE2b-256 5f39f9981ccf128a163f91486ea1e4379c12926bbef11c347450d7cd3cdec355

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp313-cp313-macosx_15_0_arm64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp312-cp312-win_amd64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp312-cp312-win_amd64.whl
Algorithm Hash digest
SHA256 687940bc00a48a4b20f781035a1e34b91d18d71889722454cca74d8df88d2e9e
MD5 09fdd5563b03f3822f67b8101170f069
BLAKE2b-256 028a3c0158bb901dbbf6956355f1378bfecdea4ac7daaafc02f529af30769a16

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp312-cp312-win_amd64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp312-cp312-manylinux_2_35_x86_64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp312-cp312-manylinux_2_35_x86_64.whl
Algorithm Hash digest
SHA256 a8100a0bdd425e2ae5da28a12832a31a9ba90ca4cad1a7ab23daa651a7e1eeaa
MD5 cd1c2bff9da80f30767899b72191ac7d
BLAKE2b-256 730b2ea052ad3c18b4890d8326a43c6a6cdac4af4f02056fc2adad548cf0b2f3

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp312-cp312-manylinux_2_35_x86_64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp312-cp312-macosx_15_0_arm64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp312-cp312-macosx_15_0_arm64.whl
Algorithm Hash digest
SHA256 e03167166644aac8f0b16bd0bd1308675ba3c4a3e47eaffa37f2bcda9cfdadce
MD5 ce9ec267291d3eb1f7569a7a942b562f
BLAKE2b-256 70991307e250e353c85c54f4078b5c64a5f00aa374c03004dcaccc7a1446147b

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp312-cp312-macosx_15_0_arm64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp311-cp311-win_amd64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp311-cp311-win_amd64.whl
Algorithm Hash digest
SHA256 f8309e4bccf5331d4347dd9ee39b245619a4ddf92e54910afe2aed88a1b20057
MD5 b4a65578434cfb10f5c4c8fe1b14f9c4
BLAKE2b-256 7e98d0d6110d4fb66f38472f92efe6ac58fbafb1e452841f925f0c64a95612be

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp311-cp311-win_amd64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp311-cp311-manylinux_2_35_x86_64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp311-cp311-manylinux_2_35_x86_64.whl
Algorithm Hash digest
SHA256 a2193e5d489b98573ad2fafb4a742c89207230bb64c3cdc6858d4a4153efe750
MD5 887d47cc508713b61fac20558596555a
BLAKE2b-256 2acb2fb7083e6408ce7e6508df0aeaa11077be57057462e384a617215a85c26a

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp311-cp311-manylinux_2_35_x86_64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file vernon_lang-0.1.1-cp311-cp311-macosx_15_0_arm64.whl.

File metadata

File hashes

Hashes for vernon_lang-0.1.1-cp311-cp311-macosx_15_0_arm64.whl
Algorithm Hash digest
SHA256 59323568b009854306fd1e363438199a23e8af7f6b2c88edad5c365750a4c0d5
MD5 6ee361e8b04e386f8c10533b70a4e57b
BLAKE2b-256 22601693bad620894d9290239a58bb7cce826989cd0ea4c33e4f974c6a6c4507

See more details on using hashes here.

Provenance

The following attestation bundles were made for vernon_lang-0.1.1-cp311-cp311-macosx_15_0_arm64.whl:

Publisher: release.yml on morphling666/VernonDSL

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.1.1 This release

12 files

0.1.0

1 file

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page