DBUS C++ bindings?

[Aristid Breitkreuz]

Am Samstag, den 25.02.2006, 11:10 +0100 schrieb Murray Cumming:
> [snip]
> > It would probably be best to stick to the "libstdc++/boost" stack
> 
> I recommend against depending on boost. It's a fine place for
> experiments, but it's not an API/ABI-stable dependency.

Not using things like boost::function is a pity, but oh well if it's not
doable, it's... not doable. The C++ standard library should be used,
however.

> 
> But it shouldn't be necessary. There won't be much objection from anyone
> to using libsigc++ if you need to.

Yep. Are signals necessary at all?

> 
> However, if the signal parameters (and return type) are decided at run
> time then neither of these compile-time type-safe signal APIs will be
> very useful.

I disagree. Something like a dbus::value_container could be used which
can save any data DBus has a notion of. However, as code usually deals
with one certain type, templates could probably come to the rescue.

PS: Feels strange, talking about vaporware.