(for upgrading users, please see the notes at the bottom)
Bazel brought a lot of nice things to the table, such as rebuilds based on
content changes instead of modification times, caching of build products,
detection of incorrect build rules via a sandbox, and so on. Rewriting the build
in Bazel was also an opportunity to improve on the Makefile-based build we had
prior, which was pretty poor: most dependencies were external or not pinned, and
the build graph was poorly defined and mostly serialized. It was not uncommon
for fresh checkouts to fail due to floating dependencies, or for things to break
when trying to switch to an older commit.
For day-to-day development, I think Bazel served us reasonably well - we could
generally switch between branches while being confident that builds would be
correct and reasonably fast, and not require full rebuilds (except on Windows,
where the lack of a sandbox and the TS rules would cause build breakages when TS
files were renamed/removed).
Bazel achieves that reliability by defining rules for each programming language
that define how source files should be turned into outputs. For the rules to
work with Bazel's sandboxing approach, they often have to reimplement or
partially bypass the standard tools that each programming language provides. The
Rust rules call Rust's compiler directly for example, instead of using Cargo,
and the Python rules extract each PyPi package into a separate folder that gets
added to sys.path.
These separate language rules allow proper declaration of inputs and outputs,
and offer some advantages such as caching of build products and fine-grained
dependency installation. But they also bring some downsides:
- The rules don't always support use-cases/platforms that the standard language
tools do, meaning they need to be patched to be used. I've had to contribute a
number of patches to the Rust, Python and JS rules to unblock various issues.
- The dependencies we use with each language sometimes make assumptions that do
not hold in Bazel, meaning they either need to be pinned or patched, or the
language rules need to be adjusted to accommodate them.
I was hopeful that after the initial setup work, things would be relatively
smooth-sailing. Unfortunately, that has not proved to be the case. Things
frequently broke when dependencies or the language rules were updated, and I
began to get frustrated at the amount of Anki development time I was instead
spending on build system upkeep. It's now about 2 years since switching to
Bazel, and I think it's time to cut losses, and switch to something else that's
a better fit.
The new build system is based on a small build tool called Ninja, and some
custom Rust code in build/. This means that to build Anki, Bazel is no longer
required, but Ninja and Rust need to be installed on your system. Python and
Node toolchains are automatically downloaded like in Bazel.
This new build system should result in faster builds in some cases:
- Because we're using cargo to build now, Rust builds are able to take advantage
of pipelining and incremental debug builds, which we didn't have with Bazel.
It's also easier to override the default linker on Linux/macOS, which can
further improve speeds.
- External Rust crates are now built with opt=1, which improves performance
of debug builds.
- Esbuild is now used to transpile TypeScript, instead of invoking the TypeScript
compiler. This results in faster builds, by deferring typechecking to test/check
time, and by allowing more work to happen in parallel.
As an example of the differences, when testing with the mold linker on Linux,
adding a new message to tags.proto (which triggers a recompile of the bulk of
the Rust and TypeScript code) results in a compile that goes from about 22s on
Bazel to about 7s in the new system. With the standard linker, it's about 9s.
Some other changes of note:
- Our Rust workspace now uses cargo-hakari to ensure all packages agree on
available features, preventing unnecessary rebuilds.
- pylib/anki is now a PEP420 implicit namespace, avoiding the need to merge
source files and generated files into a single folder for running. By telling
VSCode about the extra search path, code completion now works with generated
files without needing to symlink them into the source folder.
- qt/aqt can't use PEP420 as it's difficult to get rid of aqt/__init__.py.
Instead, the generated files are now placed in a separate _aqt package that's
added to the path.
- ts/lib is now exposed as @tslib, so the source code and generated code can be
provided under the same namespace without a merging step.
- MyPy and PyLint are now invoked once for the entire codebase.
- dprint will be used to format TypeScript/json files in the future instead of
the slower prettier (currently turned off to avoid causing conflicts). It can
automatically defer to prettier when formatting Svelte files.
- svelte-check is now used for typechecking our Svelte code, which revealed a
few typing issues that went undetected with the old system.
- The Jest unit tests now work on Windows as well.
If you're upgrading from Bazel, updated usage instructions are in docs/development.md and docs/build.md. A summary of the changes:
- please remove node_modules and .bazel
- install rustup (https://rustup.rs/)
- install rsync if not already installed (on windows, use pacman - see docs/windows.md)
- install Ninja (unzip from https://github.com/ninja-build/ninja/releases/tag/v1.11.1 and
place on your path, or from your distro/homebrew if it's 1.10+)
- update .vscode/settings.json from .vscode.dist
On a Linux machine here, the tests consistently fail when two copies
of black are run at once:
% bazel test //qt:format_check //pylib:format_check --cache_test_results=no
==================== Test output for //qt:format_check:
Process SyncManager-1:
Traceback (most recent call last):
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/process.py", line 315, in _bootstrap
self.run()
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/process.py", line 108, in run
self._target(*self._args, **self._kwargs)
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/managers.py", line 583, in _run_server
server = cls._Server(registry, address, authkey, serializer)
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/managers.py", line 156, in __init__
self.listener = Listener(address=address, backlog=16)
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/connection.py", line 453, in __init__
self._listener = SocketListener(address, family, backlog)
File "/home/dae/.cache/bazel/_bazel_dae/fc22e40cbbf8b7d16ac57a00991b1ef1/external/python/lib/python3.9/multiprocessing/connection.py", line 596, in __init__
self._socket.bind(address)
OSError: [Errno 98] Address already in use
I dug briefly into Black's code, but suspect this is actually an issue
with the multiprocessing library. Didn't have time to investigate it
further; this workaround will do for now.
(One day I'll get around to merging those separate scripts into a single
one. One day. :-))
All platforms:
- rename scripts/ to tools/: Bazelisk expects to find its wrapper script
(used by the Mac changes below) in tools/. Rather than have a separate
scripts/ and tools/, it's simpler to just move everything into tools/.
- wheel outputs and binary bundles now go into .bazel/out/dist. While
not technically Bazel build products, doing it this way ensures they get
cleaned up when 'bazel clean' is run, and it keeps them out of the source
folder.
- update to the latest Bazel
Windows changes:
- bazel.bat has been removed, and tools\setup-env.bat has been added.
Other scripts like .\run.bat will automatically call it to set up the
environment.
- because Bazel is now on the path, you can 'bazel test ...' from any
folder, instead of having to do \anki\bazel.
- the bat files can handle being called from any working directory,
so things like running "\anki\tools\python" from c:\ will work.
- build installer as part of bundling process
Mac changes:
- `arch -arch x86_64 bazel ...` will now automatically use a different
build root, so that it is cheap to switch back and forth between archs
on a new Mac.
- tools/run-qt* will now automatically use Rosetta
- disable jemalloc in Mac x86 build for now, as it won't build under
Rosetta (perhaps due to its build scripts using $host_cpu instead of
$target_cpu)
- create app bundle as part of bundling process
Linux changes:
- remove arm64 orjson workaround in Linux bundle, as without a
readily-available, relatively distro-agonstic PyQt/Qt build
we can use, the arm64 Linux bundle is of very limited usefulness.
- update Docker files for release build
- include fcitx5 in both the qt5 and qt6 bundles
- create tarballs as part of the bundling process
mypy's move to external types-* packages is a PITA, as it requires them
to be installed in site-packages, and provides no way to specify a custom
site-packages folder, necessitating extra scripts to mock the
site-packages path, and copy+rename the stub packages into a separate
folder.
Instead of generating a fluent.proto file with a giant enum, create
a .json file representing the translations that downstream consumers
can use for code generation.
This enables the generation of a separate method for each translation,
with a docstring that shows the actual text, and any required arguments
listed in the function signature.
The codebase is still using the old enum for now; updating it will need
to come in future commits, and the old enum will need to be kept
around, as add-ons are referencing it.
Other changes:
- move translation code into a separate crate
- store the translations on a per-file/module basis, which will allow
us to avoid sending 1000+ strings on each JS page load in the future
- drop the undocumented support for external .ftl files, that we weren't
using
- duplicate strings in translation files are now checked for at build
time
- fix i18n test failing when run outside Bazel
- drop slog dependency in i18n module
- anki._backend stores the protobuf files and rsbackend.py code
- pylib modules import protobuf messages directly from the
_pb2 files, and explicitly export any will be returned or consumed
by public pylib functions, so that calling code can import from pylib
- the "rsbackend" no longer imports and re-exports protobuf messages
- pylib can just consume them directly.
- move errors to errors.py
Still todo:
- rsbridge
- finishing the work on rsbackend, and check what we need to add
back to the original file location to avoid breaking add-ons
pylint fails to read Qt modules when invoked as a subprocess with
-j 0, and it looks like I committed this in the middle of debugging
the issue. Work around it by invoking pylint directly. It's still
awfully slow, taking 30 seconds on a 10 core machine.
Running and testing should be working on the three platforms, but
there's still a fair bit that needs to be done:
- Wheel building + testing in a venv still needs to be implemented.
- Python requirements still need to be compiled with piptool and pinned;
need to compile on all platforms then merge
- Cargo deps in cargo/ and rslib/ need to be cleaned up, and ideally
unified into one place
- Currently using rustls to work around openssl compilation issues
on Linux, but this will break corporate proxies with custom SSL
authorities; need to conditionally use openssl or use
https://github.com/seanmonstar/reqwest/pull/1058
- Makefiles and docs still need cleaning up
- It may make sense to reparent ts/* to the top level, as we don't
nest the other modules under a specific language.
- rspy and pylib must always be updated in lock-step, so merging
rspy into pylib as a private module would simplify things.
- Merging desktop-ftl and mobile-ftl into the core ftl would make
managing and updating translations easier.
- Obsolete scripts need removing.
- And probably more.
pytest will show what differs in simple assert statements
concurrent mode is supported with a plugin, but like nose2, concurrent
mode hides the cause of import errors, so I've left it off for now.
