On Wed, Nov 16, 2005 at 11:51:23PM +0100, Martin Dvh wrote: > Eric Blossom wrote: > >FYI, > > > >fftsink and scopesink were just refactored to use messages and message > >queues to communicate between the C++ and Python side. This removes > >our dependency on pipes, but more importantly, the C++ side will now > >never block trying to send data to the GUI.
> Thank you very much. Not depending on pipes is very important for > the windows port. I tested rightaway and fftsink and scopesink now > work right out-of-the-box on windows. (No more windows-special > hacks in scopesink and fftsink needed) Thanks for the feedback. Glad to hear it's working! > All options work, only the gui doesn't respond when the datarate is too > great. > > The C++ side will never block, but the gui does block when the data is > coming in too fast. > In fft_sink you can solve this by adding the fft_rate=5 parameter when > initiating the fft_sink. > (5 frames/sec in stead of default 20 frames/sec) > I also do this most of the time in linux because otherwise the gui-response > is very slow on my 1Ghz machine. > On windowas it just doesn't respond at all without this. This could be > different on faster machines. I think we need a .gnuradiorc kind of thing for user defaults. > The thing I couldn't built under linux was the fpga firmware because I > couldn't get quartus to work under wine. If you figure out how to get it running under wine, let us know! Eric _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
