This idea makes real sense, although probably not for all definitions, since some of them may require constant (= compile time) arguments.
Added on the TODO list. Stéphane > Le 3 févr. 2020 à 09:14, Julius Smith <julius.sm...@gmail.com> a écrit : > > It would be cool if the Faust compiler offered a special debug option, > such as "-da | --debug-all" that tries to compile every function > defined in the file separately, presenting the failed functions in > lexical order, giving line number, etc. It would be like defining > each function as process, one after the other. This is essentially > how I debug. I set process to some simpler subset of my patch, and > grow it out until it's all there. > > - Julius > > On Sun, Feb 2, 2020 at 11:42 PM Stéphane Letz <l...@grame.fr> wrote: >> >> >> >>> Le 3 févr. 2020 à 02:10, Daniel Chapiro <dchap...@gmail.com> a écrit : >>> >>> Hi Stephane, >>> Indeed, the bug was exactly where you suggested! After I had identified >>> that SR was related to the problem, how did you know it was in that >>> particular use of SR within my .dsp? >> >> By looking at : >> >> - the error message (even if cryptic ((-; ) >> >> - knowing the fact that ma.SR is defined in the maths.lib as SR = >> min(192000.0, max(1.0, fconstant(int fSamplingFreq, <math.h>))); which >> actually appears in the cryptic error message... >> >> - looking at additional constants in the cryptic error message..., and so >> finding the appropriate place in the source code. >> >> >>> >>> Also, for obscure (at least for me) error messages like the one below, is >>> there a better way than greping for some of the strings from the error msg >>> against one’s .dsp and the .libs (to narrow down possibilities), and then >>> go through all ocurrences of the likely problematic symbol until one finds >>> the bug? >>> >>> Thanks! >>> Dan >>> >> >> The error messing system is still somewhat similar to the « early days of >> template meta-programming in C++ ». The point is that precise location of >> every definitions, and the fact that the Faust compiler does a kind of « >> block diagram flattening » mainly explains the current non satisfactory (to >> say the least..) current situation. >> >> Stéphane >> >> >> >> _______________________________________________ >> Faudiostream-users mailing list >> Faudiostream-users@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/faudiostream-users > > > > -- > "Anybody who knows all about nothing knows everything" -- Leonard Susskind _______________________________________________ Faudiostream-users mailing list Faudiostream-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/faudiostream-users
