[asio-users] [http] Formal review starts on Friday

[Bjorn Reese]

This is a short notice that the formal review for the Boost.Http
library starts on Friday and will last for ten days: from August
7th until August 16th.

A proper announcement will be made on Friday, which will include
links to the source code and the documentation to be reviewed.

As a courtesy to the ASIO community we encourage your participation
in the formal review as well.

[Niall Douglas]

Post by Bjorn Reese
This is a short notice that the formal review for the Boost.Http
library starts on Friday and will last for ten days: from August
7th until August 16th.
A proper announcement will be made on Friday, which will include
links to the source code and the documentation to be reviewed.
As a courtesy to the ASIO community we encourage your participation
in the formal review as well.

You might want to explain to asio-users what the relevance of Http is
to ASIO?

Niall

[Vinícius dos Santos Oliveira]

Post by Niall Douglas
You might want to explain to asio-users what the relevance of Http is
to ASIO?

The last version of the HTTP protocol was standardized on 1999, but heavy

Post by Niall Douglas
development on the area continues to happen even today, with solutions to
increase cooperation among different applications to handle requests (WSGI,
...), new protocols created just to increase usefulness of HTTP (WebSocket)
and major new versions (currently under discussion with people dedicating
resources to eliminate problems in the drafts that could harm security,
scalability, interoperability, major adoption, ...). HTTP is used also in
some embedded devices to provide some interoperability between
applications. Even the ROS robotic framework find an use for HTTP.
In fact, there are some network connections that only allow traffic
through the ports 80 and 443 (both related to HTTP).
Boost has one of the best asynchronous I/O frameworks to C++, Boost ASIO.
It'd be really useful to have a family of protocols available on top of
Boost ASIO and HTTP is a great candidate (if not a must).

I'd add another reason today. There are many HTTP protocol implementations,
like Pion, POCO, QtHttp, Casablanca, cpp-netlib. From the mentioned ones,
Pion and cpp-netlib are the only ones who use Asio and they're limited in
their own ways[2][3]. As long as we don't provide an HTTP implementation,
we'll lose users to other async I/O frameworks. Users dealing with HTTP
want a higher abstraction than sockets.

[1] https://github.com/vinipsmaker/gsoc2014-boost#the-life-of-this-proposal
[2]
https://github.com/vinipsmaker/gsoc2014-boost/blob/master/other_frameworks.md#cpp-netlib
[3]
https://github.com/vinipsmaker/gsoc2014-boost/blob/master/other_frameworks.md#pion