libhal C++ wrapper

[Milosz Derezynski]

@David:

David, i was thinking already to remove the wrapping around libhal and
libhal-storage and basically reimplement its/their functionality using DBus
directly (basically what you said). I'm looking atm (well not _right now_
heh, but generally currently) at the libhal code, and I think this is how
hal++ is going to end up in the end.

It was a first shoot-off to replace the usage of the C API of libhal in BMP,
but i realized now too that it'd be better if i'd instead just make hal++
talk directly to hald-* through DBus instead (i guess that's what you meant
by stubs (?))

@Murray (+ David heh)

Yeah well we talked recently about dbus-cpp in #c++ and thing is i see the
whole introspectability, or rather non-introspectability, problem, as the
biggest problem, with a lot of metacode generation.

In the case of C++/gtkmm/sigc++ signaling it would be far more intrusive and
heavyweight than the C variant, as a first estimate, as e.g. signals are
hooked on to on a GSignal basis by-name: g_signal_connect (obj,
"signalname", callback, data)

With sigc++, you have to specify the signal name as a class member, etc. I
think that this is the reason no one has started real work on a dbusmm (or
continued it..) for now.

There was someone in #c++ recently (on GIMPnet) who said something about a
dbusmm/dbus-cpp/dbus-c++ but i forgot who it was sorry.

In any case the bottom line is: I'll start rewriting hal++ as something that
directly talks to hald through dbus rather than wrapping the hal libs, and
i'll just start out with the DBus C bindings, and as soon as someone comes
up with a dbusmm (or maybe i'll start on it as well or code on it with the
guy whose name i forgot :P), i can replace it internally then.

Milosz

On 11/27/06, Murray Cumming <murrayc at murrayc.com> wrote:
>
> On Mon, 2006-11-27 at 12:42 -0500, David Zeuthen wrote:
> > Did you check out dbus-cpp that Murray Cumming wrote 1-2 years ago?
> >
> >  http://webcvs.freedesktop.org/dbus/dbus-cpp/
> >
> > Perhaps it would be useful if dbus-cpp included a binding tool that
> > took
> > D-Bus introspection XML and generated stubs. Unless this is the case
> > already? Thanks.
>
> I've never done more than the very initial work on dbus C++ APIs, partly
> because I never could find a simple DBUS C API example. It should be
> better now. Basically, don't depend on me much for this.
>
> --
> Murray Cumming
> murrayc at murrayc.com
> www.murrayc.com
> www.openismus.com
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.freedesktop.org/archives/hal/attachments/20061127/fabec207/attachment.html