8.6 KiB
This — is The Construct
Internet Relay Chat daemon: Matrix Construct
IRCd is a free and open source server which facilitates real-time communication over the internet. It was started by Jarkko Oikarinen in 1988 at the University of Oulu and its derivatives underpinned the major IRC networks for decades.
In 2014 a new approach was proposed to reinvigorate real-time communication in lieu of growing proprietary competition from opaque cloud services. This is known as the Matrix Protocol: a superset of IRC that evolves it into a federation of networks and provides a means for interoperability with the modern 21st century internet messaging ecosystem.
IRCd has been rewritten to implement the Matrix Protocol using some of the latest techniques available for modern C++ free software. Just like the first iteration of IRCd, the latest Construct employs technologies in vogue for this era which provide a fulfilling experience for users and a powerfully extensible environment for developers.
Construct is designed to be fast and highly scalable, and to be community developed by volunteer contributors over the internet. This mission strives to make the software easy to understand, modify, audit, and extend. It remains true to its roots with its modular design and having minimal requirements. Even though almost all of the old code has been rewritten, the same spirit and philosophy of the predecessor is still obvious throughout.
This is the first implementation of a Matrix homeserver written in C++. The roadmap for service is as follows:
- [✓] Phase One: Matrix clients using HTTP.
- Phase Two: Legacy IRC networks using TS6 protocol (Atheme Federation).
- Phase Three: Legacy IRC clients using RFC1459/RFC2812 legacy grammars.
Installation
Getting up and running with Construct is easy. A deployment can scale from as little as a low-end virtual machine running a stock linux distribution to a large load balanced cluster operating in synchrony over a network.
Dependencies
Boost (1.61 or later) - We have replaced libratbox with the well known and actively developed Boost libraries. These are included as a submodule in this repository.
RocksDB (based on LevelDB) - We replace sqlite3 with a lightweight and embedded database and have furthered the mission of eliminating the need for external "IRC services"
Other dependencies: sodium (NaCl crypto), OpenSSL, zlib, snappy (for rocksdb)
Build dependencies: GNU C++ compiler, automake, autoconf, autoconf2.13, autoconf-archive, libtool, shtool
Notes:
- libircd requires a platform capable of loading dynamic shared objects at runtime.
Platforms
Continuously Integrated Host | Compiler | Third party | Status |
---|---|---|---|
Linux Ubuntu 16.04 Xenial | GCC 6 | Boost 1.61 |
Building from git (production)
./autogen.sh
./configure
make
sudo make install
Building from git (DEVELOPER PREVIEW INSTRUCTIONS)
This is only intended to allow development with dependencies that have not made their way to mainstream systems yet. Not for release.
The developer preview will install Construct in a specific directory isolated from the
system. It will avoid using system libraries by downloading and building the dependencies
from the submodules we have pinned here and build them the way we have configured. You may
need to set the LD_LIBRARY_PATH
to the built libraries and/or maintain an intact build
directory.
./autogen.sh
mkdir build
- The install directory may be this or another place of your choosing.
- If you decide elsewhere, make sure to change the
--prefix
in the./configure
statement below.
CXX=g++-6 ./configure --prefix=$PWD/build --enable-debug --with-included-boost=shared --with-included-rocksdb=shared
- Many systems alias
g++
to an older version. To be safe, specify a version manually inCXX
. This will also build the submodule dependencies with that version. - The
--with-included-*
will fetch, configure and build the dependencies included as submodules. Include=shared
for now until static libraries are better handled.
make
make install
Developers
- Generate doxygen using
/usr/bin/doxygen tools/doxygen.conf
the target directory is doc/html. Browse to doc/html/index.html
IRCd Library
The purpose of libircd
is to facilitate the execution of a server which
handles requests from end-users. The library hosts a set of pluggable modules
which may introduce the actual application features (or the "business logic")
of the server. These additional modules are found in the modules/
directory;
This library can be embedded by developers creating their own server or those who simply want to use the library of routines it provides.
libircd can be embedded in your application with very minimal overhead.
Linking to libircd from your executable allows you to customize and extend the
functionality of the server and have control over its execution, or, simply use
library routines provided by the library without any daemonization. Including
libircd headers will not include any other headers beyond those in the standard
library, with minimal impact on your project's compile complexity. The
prototypical embedding of libircd
is construct
found in the construct/
directory.
libircd runs only one server at a time.
Keeping with the spirit of simplicity of the original architecture, libircd
continues to be a "singleton" object which uses globals and keeps actual server
state in the library itself. In other words, only one IRC daemon can exist
within a process's address space at a time. Whether or not this was a pitfall
of the original design, it has emerged over the decades as a very profitable
decision for making IRCd an accessible open source internet project.
libircd is single-threaded✝
The library is based around the boost::asio::io_service
event loop. It is
still an asynchronous event-based system. We process one event at a time;
developers must not block execution. While the io_service
can be run safely
on multiple threads by the embedder's application, libircd will use a single
io_service::strand
.
This methodology ensures there is an uninterrupted execution working through a single event queue providing service. If there are periods of execution which are computationally intense like parsing, hashing, cryptography, etc: this is absorbed in lieu of thread synchronization and bus contention. Scaling this system is done through running multiple instances which synchronize at the application level.
✝ However, don't start assuming a truly threadless execution for the entire
address space. If there is ever a long-running background computation or a call
to a 3rd party library which will do IO and block the event loop, we may use an
additional std::thread
to "offload" such an operation. Thus we do have
a threading model, but it is heterogeneous.
libircd introduces userspace threading✝
IRCd presents an interface introducing stackful coroutines, a.k.a. userspace context switching, or green threads. The library avoids callbacks as the way to break up execution when waiting for events. Instead, we harken back to the simple old ways of synchronous programming where control flow and data are easy to follow.
✝ If there are certain cases where we don't want a stack to linger which may jeopardize the c10k'ness of the daemon the asynchronous pattern is still used.
libircd innovates with formal grammars
We leverage the boost::spirit
system of parsing and printing through formal grammars,
rather than writing our own parsers manually. In addition, we build several tools
on top of such formal devices like a type-safe format string library acting as a
drop-in for ::sprintf()
, but accepting objects like std::string
without .c_str()
and prevention of outputting unprintable/unwanted characters that may have been
injected into the system somewhere prior.