Re: Further gstreamermm wrapping

[Murray Cumming <murrayc murrayc com>]

On Fri, 2008-03-21 at 23:36 +0100, Milosz Derezynski wrote:
> There is one difference between the rest of the plugins as found in
> the -good, -bad and -ugly packages (and 3rd party plugins): Those are
> the base types which everything derives of.
> 
> The reason why it would make sense to explicitly wrap the core and
> base plugins (e.g. GstBaseSrc) is that then it would be possible to
> write wholly new elements (plugins) using our C++ binding;

Ah, OK. Then these base plugin types should definitely be wrapped.

>  otherwise this would be not possible, not even with the code
> generator i'm working on, as this only wraps properties and signals,
> which is enough for normal elements as they usually don't expose
> specific API in the sense of methods, they only have own properties
> and signals to be attached to.
> 
> "Normal" here refers to anything that isn't a base for anything else.
> Derivation from specific elements isn't possible or needed either (and
> hence it's not neccessary to wrap all of them which would be a
> daunting task to keep up with, if possible at all), because they
> usually don't extend the object structure beyond what they've
> inherited from their base (e.g. GstGIOSrc only fulfills the GstBaseSrc
> API, but doesn't extend it).
-- 
Murray Cumming
murrayc murrayc com
www.murrayc.com
www.openismus.com