rap@dana.UUCP (06/19/87)
I have sent to amiga-sources-request the audio material from the Amiga World article. I've also sent a somewhat more documented version to FISH. If you want the enhanced version, I am still working on it. I expected the July issue to be hitting the stands in July. So what did I know? Somewhere after July 1 I'll have some of the enhancements done, and perhaps other suggestions in hand. I already figure to try to integrate the speech with the audio, since there are, after all, four channels to allocate and fool around with. Only problem is that speech takes up so very much of the CPU time. Ah well. Enjoy (and by all means, let me know if you find any bugs.) Rob Peck. ...hplabs!dana!rap
rap@dana.UUCP (Rob Peck) (06/22/87)
In article <182@dana.UUCP>, rap@dana.UUCP (Rob Peck) writes: > I have sent to amiga-sources-request the audio material from the > Amiga World article. > > Enjoy (and by all means, let me know if you find any bugs.) > > Rob Peck. ...hplabs!dana!rap > Found a bug myself, final version of fix not available yet, but I do know where the bug itself is. Bug is in FinishAudio(). ... if FinishAudio gets called before the audio device has returned everything to be reused, the program can hang waiting for something that will never happen. FinishAudio now has no way to go back and wait for everything to finish before it exits. It checks whether everything is done, but does not do the job completely. The fix is to keep track of how many audio requests are outstanding for all channels combined, and go back to waiting on the reply port and emptying the port, decrementing the count, until the count on all channels is zero. Will post the fix as soon as I have it available. Just thought I'd let everyone know (found problem while creating PlayFreq, queued up 400 freqs, and decreased the Delay value; static iob got reused, FinishAudio got called before dynamics all got freed, showed the problem.). Rob Peck. ...hplabs!dana!rap