Re: [Dbus-cxx-users] Maintenance

[Robert M. <rmi...@sy...>]

Not sure if you got my last e-mail, but yes if you could add me to the
project I can commit what changes we have.  E-mail me directly.

Quick list of bugfixes that we've done:
-hang on method calls that return void(possibly fixed, I noticed an issue
with it the other day)
-Updated dispatcher to continue on if select() fails because of an
interrupted system call(apparently this happens when the program receives a
signal and it's handled)
-forced use of C++0x smart pointers as boost would throw errors on our
system.  May be a problem with the boost library version.
-If you remove a signal callback, you won't receive that signal
anymore(your program would never receive the signal again, but the D-Bus
would still send out the signal to you)
-changed PendingCall to not callback if there are no more references to
it(this could occur if you used dbus_connecton->send_with_reply_async() and
didn't save the pointer that you got back - your callback would be (most
likely) deconstructed by the time the D-Bus returned a reply)
-added variants
-fixed a bug where a segfault would occur if the interface that was called
was NULL
-some better error handling mechanisms so that we won't exit if something
bad happens.  This has to do with calling remote methods when an error is
thrown from the remote method.

-Robert Middleton

On Wed, Feb 22, 2012 at 5:31 PM, Rick L. Vinyard, Jr.
<rvi...@cs...>wrote:

> Just haven't had time.
>
> I'll add you to the project and give you access to the repo if you want.
>
> Robert Middleton wrote:
> > I noticed yesterday that somebody had posted a message asking if dbus-cxx
> > was still being maintained.  Rick said that he'd commit stuff that was
> > tested, so what's the deal for that stuff to be tested?  We've been using
> > dbus-cxx for well over a year now, and we've fixed quite a few bugs in
> > that
> > time(we posted on the dev lists a year ago regarding Variants).
> >
> > -Robert Middleton
> >
>
>
>