As of this writing, Cap‘n Proto is in beta. The serialization layer is close to feature-complete and we don’t anticipate any further changes to the wire format. That said, if you want to use it, you should keep in mind some caveats:
If you‘d like to hack on Cap’n Proto, you should join the discussion group!
If you'd just like to receive updates as things progress, add yourself to the announce list.
The Cap'n Proto tools, including the compiler which takes .capnp files and generates source code for them, are written in C++. Therefore, you must install the C++ package even if your actual development language is something else.
If you are using GCC, you MUST use at least version 4.7 as Cap‘n Proto uses recently-implemented C++11 features. If GCC 4.7 is installed but your system’s default GCC is older, you will probably need to set the environment variable CXX=g++-4.7 before following the instructions below.
If you are using Clang, you must use at least version 3.2. To use Clang, set the environment variable CXX=clang++ before following any instructions below, otherwise g++ is used by default.
This package is officially tested on Linux (GCC 4.7, Clang 3.2), Mac OSX (Clang 3.2), and Cygwin (Windows; GCC 4.7), in 32-bit and 64-bit modes.
Mac OSX users: Don't miss the special instructions for OSX.
You may download and install the release version of Cap'n Proto like so:
This will install capnp, the Cap'n Proto command-line tool. It will also install libcapnp, libcapnpc, and libkj in /usr/local/lib and headers in /usr/local/include/capnp and /usr/local/include/kj.
If you download directly from Git, and you don't want to build with Ekam, you will need to have the GNU autotools -- autoconf, automake, and libtool -- installed. You will also need Subversion installed (in addition to Git) in order to fetch the Google Test sources (done by setup-autotools.sh).
git clone https://github.com/kentonv/capnproto.git cd capnproto/c++ ./setup-autotools.sh autoreconf -i ./configure make -j6 check sudo make install
As of this writing, Mac OSX 10.8 with Xcode 4.6 command-line tools is not quite good enough to compile Cap‘n Proto. The included version of GCC is ancient. The included version of Clang -- which mysteriously advertises itself as version 4.2 -- was actually cut from LLVM SVN somewhere between versions 3.1 and 3.2; it is not sufficient to build Cap’n Proto.
There are two options:
Option 2 is the one preferred by Cap‘n Proto’s developers. Here are step-by-step instructions for setting this up:
Get the Xcode command-line tools: Download Xcode from the app store. Then, open Xcode, go to Xcode menu > Preferences > Downloads, and choose to install “Command Line Tools”.
Download the Clang 3.2 binaries and put them somewhere easy to remember:
curl -O http://llvm.org/releases/3.2/clang+llvm-3.2-x86_64-apple-darwin11.tar.gz tar zxf clang+llvm-3.2-x86_64-apple-darwin11.tar.gz mv clang+llvm-3.2-x86_64-apple-darwin11 ~/clang-3.2
We will need to use libc++ (from LLVM) rather than libstdc++ (from GNU) because Xcode‘s libstdc++ (like its GCC) is too old. In order for your freshly-downloaded Clang binaries to be able to find it, you’ll need to symlink it into the Clang tree:
ln -s /usr/lib/c++ ~/clang-3.2/lib/c++
You may now follow the instructions below, but make sure to tell configure to use your newly-downloaded Clang binary:
./configure CXX=$HOME/clang-3.2/bin/clang++
Hopefully, Xcode 5.0 will be released soon with a newer Clang, making this extra work unnecessary.
Ekam is a build system I wrote a while back that automatically figures out how to build your C++ code without instructions. It also supports continuous builds, where it watches the filesystem for changes (via inotify) and immediately rebuilds as necessary. Instant feedback is key to productivity, so I really like using Ekam.
Unfortunately it's very much unfinished. It works (for me), but it is quirky and rough around the edges. It only works on Linux, and is best used together with Eclipse. If you find it unacceptable, scroll up to the Automake instructions.
The Cap'n Proto repo includes a script which will attempt to set up Ekam for you.
git clone https://github.com/kentonv/capnproto.git cd capnproto/c++ ./setup-ekam.sh
If all goes well, this downloads the Ekam code into a directory called .ekam and adds some symlinks under src. It also imports the Google Test and Protobuf source code, so you can compile tests and benchmarks.
Once Ekam is installed, you can do:
make -f Makefile.ekam once
This will build everything it can and run tests.
Note that Ekam will fail to build some things and output a bunch of error messages. You should be able to ignore any errors that originate outside of the capnp and kj directories -- these are just parts of other packages like Google Test that Ekam doesn‘t fully know how to build, but aren’t needed by Cap'n Proto anyway.
Before getting into benchmarks, let me be frank: performance varies wildly by use case, and no benchmark is going to properly reflect the big picture. If performance is critical to your use case, you should write a benchmark specific to your case, and test multiple serialization technologies. Don‘t assume anything. If you find Cap’n Proto performs sub-optimally, though, tell us about it.
That said, Cap'n Proto does have a small suite of silly benchmarks used to validate changes.
The Ekam build will put the benchmark binaries in tmp/capnp/benchmark.
tmp/capnp/benchmark/runner
This runs the default test case, CatRank. CatRank simulates a search engine scoring algorithm which promotes pages that discuss cats (and demotes ones discussing dogs). A list of up to 1000 random search results with URLs, scores, and snippets is sent to the server, which searches the snippets for instances of “cat” and “dog”, adjusts their scores accordingly, then returns the new result list sorted by score.
This test case is very string-heavy. Cap‘n Proto performs well due to its zero-copy strings, but packing the message doesn’t help much.
tmp/capnp/benchmark/runner eval
In this test case, the client generates a random, deeply-nested arithmetic expression for the server to evaluate. This case is a pathologically bad case for Cap‘n Proto as it involves lots of pointers with relatively little actual data. When packing is enabled it actually loses to Protobufs by a little bit on CPU time (as of this writing, at least; it’ll probably get better with optimization).
tmp/capnp/benchmark/runner carsales
This test case involves sending to the server a description of a bunch of cars, and asks the server to decide how much the lot is worth. This case is very number-heavy, and because of this Cap‘n Proto’s “packed” mode really shines.
If you intend to do some development, you should build continuous or continuous-opt instead of once. These modes will build everything, then watch the source tree for changes and rebuild as necessary. continuous does a debug build while continuous-opt optimizes; the former is best while developing but don't run the benchmarks in debug mode!
If you use Eclipse, you should use the Ekam Eclipse plugin to get build results fed back into your editor while building in continuous mode. Build the plugin like so:
.ekam/eclipse directory as an Eclipse project.The dashboard view lets you browse the whole tree of build actions and also populates your editor with error markers.