I do not think the amount of space for arguments should make a difference at performance time - if it seems to do for you, it may be the result of random variation and inaccuracy in the measured times, or minor differences due to data alignment etc. Where it does make a difference is memory usage and the time needed for note allocations, and there more is worse (i.e. more space wasted, longer time to clear the allocated memory to zero, etc.). It may not be a large difference, but it can add up if large numbers of opcodes are used. Of course, increasing the limit by a small amount is not a real solution anyway: what if someone needs 65 outlets ? On Friday 05 May 2006 16:57, David Akbari wrote: > For the record, I tried this solution and it works however I reverted > any changes made to the local source tree on this machine simply out of > fear that I would make something useful but it would not be portable > with subsequent canonical revisions of the language. > > After some tests to see exactly what the magnitude is for the statics > on orchestra compilation / score sort times, it seems that the current > limit (23 outargs) is actually compiling things a little slower than a > power of two value - I tried 16, 32, 48, 64. > > 16 and 64 resulted in the fastest compile times. > > If other users would find this useful and could confirm this behaviour, > then I think it might be useful to extend the maximum number of outargs > in the UDO mechanism to 64. It might also be preferable to come up with > a better solution to the statics in general but I know that this is a > non-trivial endeavor. > > As it is now, I'm leaving the limit of 23 until a more official > decision is made and using an instr with zak k-space. > > Thanks again for your help ! ------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Csound-devel mailing list Csound-devel@lists.sourceforge.net