On Monday 17 December 2001 19.36, Paul Davis wrote: > >I would be delighted if someone could tell me what the name semantics are. > > there are none. just use the string provided the user, or "default" if > they don't provide any (or "hw:0,0" if you prefer and have good > reasons for doing so).
I was suspecting that this was the trick, just providing the string, so I will do so, let the user specify the string and then just pass it along. using "hw:x,x,x" will however be the recommended mode of operaton. > >In BruteFIR I only want to be as close to the hardware as possible, > > with all due respect, since not a line of your code has to be changed > to be close to or far from the hardware, shouldn't this be a choice > offered to the user, rather than a decision made by you? Yep, you are probably right. Why it is like it is now is mostly because I am ignorant when it comes to ALSA coding. When it started working, I stopped coding it. However, BruteFIR has functions which expect to be in the last output stage before the hardware, so hindering software access is not entirely dumb. But, then again, this is Linux, not Windows, the user should be allowed to do what he or she likes, the user knows the best, not the program. > if i want to use BruteFIR on 4 channels of my 26 channel hammerfall, > why should you stop me from doing that? i understand that it may not > work as well as with "hw:dev,subdev", but thats not your fault, thats > mine for requesting this. Well, you are not stopped even now, you open 26 channels but use only four of them, you can of course specify exactly which four you wish to use, and how to order them. BruteFIR has features like dithering and peak limiter (ok, the peak limiter is just a clipper so far, but I'm working on it, sort of) and features like that should be at the very last output stage to work as expected. This makes it not reccomended to pass BruteFIR output through another software layer with mixing and stuff (for instance ALSA-lib software magic) before reaching the hardware. A recommendation to use "hw:x,x,x" in the documentation and an explanation why should be enough though. > your current approach to calling snd_pcm_open(), prohibits this, even > though the rest of your code allows it. I will fix it tonight. Anyone who knows if that will solve Gerard's problem with accessing the real channels of the the SB-Live? Perhaps that requires more changes than just passing the open string? /Anders Torger _______________________________________________ Alsa-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/alsa-devel
