ninja VS Cap'n Proto

Ninja showed real promise for a while, but then CMake grew up and people stopped seeing a reason to leave it behind.

warning: Starting with the September 2023 release, the default triplet for vcpkg libraries will change from x86-windows to the detected host triplet (x64-windows). To resolve this message, add --triplet x86-windows to keep the same behavior. Computing installation plan... The following packages will be built and installed: * egl-registry:x86-windows -> 2022-09-20 glew:x86-windows -> 2.2.0#3 * opengl:x86-windows -> 2022-12-04#3 * opengl-registry:x86-windows -> 2022-09-29#1 * vcpkg-cmake:x64-windows -> 2023-05-04 * vcpkg-cmake-config:x64-windows -> 2022-02-06#1 Additional packages (*) will be modified to complete this operation. Detecting compiler hash for triplet x86-windows... A suitable version of powershell-core was not found (required v7.2.11) Downloading portable powershell-core 7.2.11... Downloading powershell-core... https://github.com/PowerShell/PowerShell/releases/download/v7.2.11/PowerShell-7.2.11-win-x86.zip->C:\vcpkg\downloads\PowerShell-7.2.11-win-x86.zip Downloading https://github.com/PowerShell/PowerShell/releases/download/v7.2.11/PowerShell-7.2.11-win-x86.zip Extracting powershell-core... error: while detecting compiler information: The log file content at "C:\vcpkg\buildtrees\detect_compiler\stdout-x86-windows.log" is: -- Downloading https://github.com/ninja-build/ninja/releases/download/v1.10.2/ninja-win.zip -> ninja-win-1.10.2.zip... -- Configuring x86-windows CMake Error at scripts/cmake/vcpkg_execute_required_process.cmake:112 (message): Command failed: C:/vcpkg/downloads/tools/ninja/1.10.2-windows/ninja.exe -v Working Directory: C:/vcpkg/buildtrees/detect_compiler/x86-windows-rel/vcpkg-parallel-configure Error code: 1 See logs for more information: C:\vcpkg\buildtrees\detect_compiler\config-x86-windows-rel-CMakeCache.txt.log C:\vcpkg\buildtrees\detect_compiler\config-x86-windows-out.log

I thought I would share my custom installer script for the latest GitHub versions of CMake, Ninja, and Meson.

Now that you have your build system all generated you can go ahead and build everything. By default Meson will use Ninja as the build tool. Ninja is similar to Make but much much faster. You can also generate additional build systems but that's outside of the scope of this post.

There are also some other tools like https://ninja-build.org/ that you might prefer using instead

Oh excellent, then better (and more portable!) tools are available:

http://pants.build

https://ninja-build.org

https://buck.build

and, if you hate yourself: https://bazel.build

Doctor summary (to see all details, run flutter doctor -v): [✓] Flutter (Channel stable, 3.3.10, on Artix Linux 6.0.12-artix1-1, locale en_US.UTF-8) [✓] Android toolchain - develop for Android devices (Android SDK version 33.0.1) [✗] Chrome - develop for the web (Cannot find Chrome executable at google-chrome) ! Cannot find Chrome. Try setting CHROME_EXECUTABLE to a Chrome executable. [✗] Linux toolchain - develop for Linux desktop ✗ clang++ is required for Linux development. It is likely available from your distribution (e.g.: apt install clang), or can be downloaded from https://releases.llvm.org/ ✗ CMake is required for Linux development. It is likely available from your distribution (e.g.: apt install cmake), or can be downloaded from https://cmake.org/download/ ✗ ninja is required for Linux development. It is likely available from your distribution (e.g.: apt install ninja-build), or can be downloaded from https://github.com/ninja-build/ninja/releases ✗ pkg-config is required for Linux development. It is likely available from your distribution (e.g.: apt install pkg-config), or can be downloaded from https://www.freedesktop.org/wiki/Software/pkg-config/ [✓] Android Studio (version 2021.3) [✓] Connected device (1 available) [✓] HTTP Host Availability ! Doctor found issues in 2 categories.

- may massively reduce the latency involved.

Those sharing Cap'n Proto-encoded data may have particular interest. Cap'n Proto (https://capnproto.org) is fantastic at its core task - in-place serialization with zero-copy - and we wanted to make the IPC (inter-process communication) involving capnp-serialized messages be zero-copy, end-to-end.

That said, we paid equal attention to other varieties of payload; it's not limited to capnp-encoded messages. For example there is painless (<-- I hope!) zero-copy transmission of arbitrary combinations of STL-compliant native C++ data structures.

To help determine whether Flow-IPC is relevant to you we wrote an intro blog post. It works through an example, summarizes the available features, and has some performance results. https://www.linode.com/blog/open-source/flow-ipc-introductio...

Of course there's nothing wrong with going straight to the GitHub link and getting into the README and docs.

Currently Flow-IPC is for Linux. (macOS/ARM64 and Windows support could follow soon, depending on demand/contributions.)

The second one is to encode data in such a way that you can read it and operate on it directly from the buffer. You write data in a layout that is the same, or easily transformed as types in memory. To do that you usually need to encode with a known schema, only Sized types to efficiently compute fields locations as offsets in the buffer, and you usually represent pointers as offset into the encode. You can look at capnproto protocol for instance https://capnproto.org/

Worked well for Cap'n Proto (the cerealization protocol)! https://capnproto.org/

With all due respect, you read completely wrong.

* The very first use case for which Cap'n Proto was designed was to be the protocol that Sandstorm.io used to talk between sandbox and supervisor -- an explicitly adversarial security scenario.

* The documentation explicitly calls out how implementations should manage resource exhaustion problems like deep recursion depth (stack overflow risk).

* The implementation has been fuzz-tested multiple ways, including as part of Google's oss-fuzz.

* When there are security bugs, I issue advisories like this:

https://github.com/capnproto/capnproto/tree/v2/security-advi...

* The primary aim of the entire project is to be a Capability-Based Security RPC protocol.

I like how they use capability-based security [0] and use Cap'n Proto protocol. This is another technology that is slow to get broad adoption, but has many things going for when compared to e.g. Protocol Buffers (Cap'n Proto is created by the primary author of Protobuf v2, Kenton Varda).

[0] https://sandstorm.io/how-it-works#capabilities

[1] https://capnproto.org

Related but not Rust-specific: FlatBuffers, Cap'n Proto.

Cap'n proto would be ideal for this. It can read out the data from a const byte array (using include_bytes! macro) directly without a separate deserialise step. protobuf and bincode would require deserialising the data into a heap allocated structure on startup or first use. Because your binary will be memory-mapped by the OS, the data can also be lazily paged in as it is read, requiring less I/O if you don't read the whole file.

This is only possible when you don't need serialization, though. However, just last week I started looking into Cap'n Proto to solve this issue. It's a serialization format that's designed for shared memory and so provides data types that can be used directly with no conversion. There's a full Rust implementation for it.

Google thinks gRPC is fast enough for there use case. What do you consider high-performance? Plain UDP with custom TCP-like stack? https://capnproto.org/?

Is Cap'n'Proto [0] as awesome as it sounds? Anyone have any first-hand experience and can speak on it?

Thanks!

[0]: https://capnproto.org/