Hi Albert, Concerning the first part, it seems you did not read my mail of June 17, I put it again here :
======== Since we don’t want to cause any « cardiac arrest » to anybody anymore… ((-; we have created a new « remorked_dsp » branch with some work in the base dsp class. https://sourceforge.net/p/faudiostream/code/ci/reworked_dsp/ Two different things have been done : 1) a new « virtual dsp* clone() » method has been added in the base dsp class, and is now added by the compiler in the generated « mydsp » class. This method has become necessary with the new « poly » dsp model (to easily duplicate the DSP voice used in the polyphonic object) and will be needed with future extensions we plan in a more « macro construction » model of DSP construction. 2) the « static void metadata(Meta* m) » method of the dsp class has been made an object method. Defining a static method was really a C++ like way to do, and was finally more annoying that helpful in a lot of cases : C, LLCM, asm.js…etc… backends in the faust2 branch with this static model dos not make real sense..etc. The 1) change should not break anything. I did necessary changes for 2) everywhere it should be needed. I even « blindly » edited some architecture files that I cannot test : pure.cpp, vsp.cpp, dssi.cpp, supercollider.cpp, lv2ui.cpp, lv2.cpp, faustvst.cpp. This would be great is someone could test they are properly modified, before we merge the branch back in master. ======================= ==> after tests, we finally merged this branch on master then faust2. Concernant division by zero this that done by this commit b19223d88afb2b8f26226bf2453c5ead66abee1c "Detect division by zero during symbolic division », maybe the fix is not correct? Yann? Stéphane > Le 25 juin 2016 à 01:59, Albert Graef <aggr...@gmail.com> a écrit : > > I've got some trouble with the latest faust again (using the faust2 branch). > > First, it seems that metadata() was recently changed from a class to an > object method, breaking the API. Conceptually, that doesn't make much sense > to me; the dsp meta-data should be the same for all dsp instances and hence > static, no? So why was this change needed? > > Second, glassHarmonica.dsp suddenly refuses to compile, faust bails out with > this error message: > > ERROR: division by 0 in min(1.92e+05f, max(1.0f, fSamplingFreq)) * > nentry(freq [1][unit:Hz] [tooltip:Tone > frequency]/Basic_Parameters,4.4e+02f,2e+01f,2e+04f,1.0f)**-1 / 0.0f > > Looking at glassHarmonica.dsp, I don't see any division by zero there. Can > anyone enlighten me? > > TIA, > Albert > > -- > Dr. Albert Gr"af > Computer Music Research Group, JGU Mainz, Germany > Email: aggr...@gmail.com > WWW: https://plus.google.com/+AlbertGraef > ------------------------------------------------------------------------------ > Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San > Francisco, CA to explore cutting-edge tech and listen to tech luminaries > present their vision of the future. This family event has something for > everyone, including kids. Get more information and register today. > http://sdm.link/attshape_______________________________________________ > Faudiostream-devel mailing list > Faudiostream-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/faudiostream-devel ------------------------------------------------------------------------------ Attend Shape: An AT&T Tech Expo July 15-16. Meet us at AT&T Park in San Francisco, CA to explore cutting-edge tech and listen to tech luminaries present their vision of the future. This family event has something for everyone, including kids. Get more information and register today. http://sdm.link/attshape _______________________________________________ Faudiostream-devel mailing list Faudiostream-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/faudiostream-devel
