On Wed, Sep 01, 2004 at 07:41:53PM +0200, Francis ANDRE wrote:
IMHO, a RMI C++ package should be independent of the underlying transport protocol. This last one couldbe abstracted as an iostream like the fstream classes
for tcp/udp, one could use a socketstream (SocketC++ or others as starting point) for x25, one could use x25stream for pipe/named pipe, one could use pstream (@sourceforge)
For the record, that's pstreams.sf.net with an 'S', not pstream.sf.net (the latter is a java thingy of some sort) jon (pstreams author)
for file, one could use fstream
and so on
templates are good for this kind of job
FA
-- A good friend will come bail you out of jail.......... but, a true friend....will be sitting next to you saying: "...that was fun."
"Matt Hurd" <matt.hurd@gmail.com> a �crit dans le message de news:8f48d4a04083123442a8464a4@mail.gmail.com...
Francis ANDRE <francis.andre@easynet.fr> wrote:
Now that the serialization library is part of boost, could this be a item to add the the wishlist at boost.org???
tcp/udp wrapping should probably come first.
matt.