Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Murray Cumming
On Tue, 2009-12-22 at 10:39 -0600, Jonathon Jongsma wrote: On Tue, 2009-12-22 at 17:22 +0100, Krzesimir Nowak wrote: On Tue, 2009-12-22 at 15:58 +0100, Glus Xof wrote: 2009/12/22 Krzesimir Nowak qdl...@gmail.com: For sockets - use GSocket. But C++ wrapper of GSocket just hit the

Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Jonathon Jongsma
On Fri, 2010-01-29 at 15:39 +0100, Murray Cumming wrote: Speaking of which: I would really appreciate people trying it out, testing, and/or helping finish it. My time has been rather short lately. In particular, there is one example program using Gio::Resolver checked into git

Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Murray Cumming
On Fri, 2010-01-29 at 08:56 -0600, Jonathon Jongsma wrote: Oh, I forgot to follow up about this problem. I have fixed this particular issue (it was a function that wasn't returning a ref on an object and we forgot to take an extra ref). There are still a few methods here and there that need

Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Jonathon Jongsma
On Fri, 2010-01-29 at 16:11 +0100, Murray Cumming wrote: On Fri, 2010-01-29 at 08:56 -0600, Jonathon Jongsma wrote: Oh, I forgot to follow up about this problem. I have fixed this particular issue (it was a function that wasn't returning a ref on an object and we forgot to take an extra

Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Krzysztof Kosiński
2010/1/29 Jonathon Jongsma jonat...@quotidian.org: But nearly every function in gio has an overload for when exceptions are disabled, so... Maybe glibmm 3.0 should just require exceptions and remove the no-exception mode? Is the exception handling overhead large enough to warrant maintaining

Re: Glib::IOChannel in sockets... (2)

2010-01-29 Thread Jonathon Jongsma
On Fri, 2010-01-29 at 20:08 +0100, Krzysztof Kosiński wrote: 2010/1/29 Jonathon Jongsma jonat...@quotidian.org: But nearly every function in gio has an overload for when exceptions are disabled, so... Maybe glibmm 3.0 should just require exceptions and remove the no-exception mode? Is the