Hi folks, I've recently checked in a bunch of changes, some of which were to gr_block.h, which everything depends on.
As a result, if you are tracking CVS, you will need to update and rebuild *everything*. One of the changes altered how we deal with the history mechanism. This is mostly transparent, but filters now get a zero initialized delay line, and produce N outputs when N inputs are run through them. The old code used to produce N-ntaps-1 outputs when N inputs were run through them. These changes allow transmitted data to be sent discontinuously to the USRP. The additional layer of transmit buffering that used to be in the fusb_linux code has been removed. By changing the default number of samples/baud from 4 to 2, the GMSK code is now running at 1Mbit/sec with the same error performance ;) This is without any real tuning... 2Mbit/sec should definitely be doable. Regarding 2.6: I still need to clean up the boost autoconf code and confirm that "make distcheck" works on our favorite platforms. Other than that, we're good to go. Eric _______________________________________________ Discuss-gnuradio mailing list Discuss-gnuradio@gnu.org http://lists.gnu.org/mailman/listinfo/discuss-gnuradio
