10.02.2015 01:57, David Henningsson wrote: > Thanks for the review! > > On 2015-02-01 15:37, Alexander E. Patrakov wrote: >> If adding proper comments to all the math is a too-hard requirement, >> then, in v2, I'd suggest to (temporarily?) squash patches 1 and 2 for >> easier review, and to remove all unused functions and parameters. > > So to answer to these more general thoughts: > > - Yes, I think removing dead code makes sense. I was thinking that > maybe it makes sense to have it in case we wanted to implement more > filters later, but perhaps we should take that problem when it happens. > > - Adding comments and references is going to be difficult as I'm not > an DSP expert. > > - As for squashing patches, I think it's important that patch 1/6 > remains the way it is for tracking/Copyright purposes (i e who wrote > what code). I can commit to that in v2, I'll make three patches - the > two first ones and a third to remove dead code - and also post the > combined diff of these three for easier reviewing. Well, I think that, after removal of dead code (which reduces the amount of work), copyrights can be made a non-issue. After all, you can always reimplement the coefficient calculation from scratch according to this cookbook: http://www.musicdsp.org/files/Audio-EQ-Cookbook.txt Then compare the result of your own implementation with those from the webkit-copyrighted one, and, if they match, remove the webkit-copyrighted one. -- Alexander E. Patrakov
