Re: [sc-users] scplug (was: hobnox)

hi dan,

On 15.05.2008, at 22:46, Dan Stowell wrote:
> I would really like to be able to understand this code so I can maybe
> play with it in meaningful ways. I can't currently compile it (can't
> install mozilla-dev ATM) but could you help with some pointers please?

i sat down today and started porting to the qtbrowserplugin framework,
so hopefully in a few days it's functional (again). with the qt stuff it
should run and compile on all three platforms SC runs on and be usable
in all major browsers -- except for IE, which is only supported by the
commercial version of qt.

> - The np*.cpp files - there are 3 of these, what do they each do? I'm
> sure they connect to the NPAPI, but I wonder what their 3 different
> roles are. (In fact, the same goes for the other c++ files...)

those files were copied from an example plugin and most of the code is
boilerplate for providing a c++ layer to the low-level c interface. the
real thing is in SC_NPPlugin.(h|cpp)

> - The readme says "currently Linux only" - what's the reason for
> that? Just a makefile thing, or...?

obtaining a mozilla build setup (especially with the IDL compiler etc)
on anything else than debian seems to be a pita, this is going to be
much smoother with qtbrowserplugin.

> - Is OSC used for communication between browser and plugin? I'm a bit
> confused - the supercollider.js looks like it's got objects for
> handling OSC but I can't see where the JS for actually
> generating/receiving the OSC would be.

the plugin links to the internal server and communication is via OSC
through a message fifo. the JS code calls the sendMessage and sendBundle
plugin methods, which marshal the argument JS objects to OSC packets and
push them onto the fifo. reply messages from the synth are also pulled
from a message queue -- unfortunately this queue has to be polled from a
JS timer currently, hopefully there'll be a way in the future to
directly call back into the JS interpreter.

thanks for your interest! as soon as the switch to qt is done we'll have
a more solid basis for further development. i'll keep you posted ...

<sk>

_______________________________________________
Sc-devel mailing list
Sc-devel@...
http://lists.create.ucsb.edu/mailman/listinfo/sc-devel