Boost logo

Boost :

From: Jeff Garland (jeff_at_[hidden])
Date: 2005-07-14 23:01:32


On Thu, 14 Jul 2005 17:40:34 -0600, Jonathan Turkanis wrote
> > 1. Inter-process communications, including simple popen-type usage,
> > like a scripting language; also pipes for forking
> > 2. Socket communications, TCP UDP and unix
>
> For the 1.34 release of Boost, I'm planning to add socket and pipe
> Devices to Boost.Iostreams, which is debuting in 1.33
> (http://boost-consulting.com/boost/libs/iostreams)

There's also been some work on Boost.Process -- see:

http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl?BoostProcess

But it's not official yet.

> > 3. Serial communications, and a wrapper to set the serial parameters
> >
> > For 1. and 2., we are currently using the socket++ library. I am also
> > aware of the PStreams library for 1.
> > for 3. we use hacked together mix of C and C++.
> >
> > Is there any sort of joint strategy planned to tackle this area? If
> > there is, I may be able to contribute.
>
> There's been a lot of work on networking, most recently by Pedro
> Lamarão (see the archives). I'm sure your contributions would be welcomed.

And formally by Hugo Duncan and many others:

http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl?BoostSocket
http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl?Multiplexing

I've thought for some time now we need a collaborative team to work the
networking area. Reason being that we never seem to get a single author that
can carry it end to end (not enough Robert Ramey's around :). A motiviated
group might be able to make progress much faster and finally get a library in
boost. Nicholai Josuttis and I had some conversations at OOPSLA last year with
'an acedemic person famous for C++ networking libraries' and he indicated an
interest in collaborating on getting a networking library for boost. We
haven't followed up on this offer and we really should. If you are highly
motiviated, contact me offlist and maybe we can start something....

BTW, the above is not meant to discourage efforts by Pedro, Jonathan, and
others. It's just that we've seen various efforts start and die over many
years. We really need a proposal for TR2, and well time is slipping away...

Jeff


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk