Date: Thu, 16 Oct 1997 19:44:18 -0400 From: Randall Hopper <rhh@ct.picker.com> To: Luigi Rizzo <luigi@labinfo.iet.unipi.it> Cc: multimedia@FreeBSD.ORG Subject: Re: Who is interested in /dev/midi and /dev/synth ? Message-ID: <19971016194418.46875@ct.picker.com> In-Reply-To: <199710160624.HAA10739@labinfo.iet.unipi.it>; from Luigi Rizzo on Thu, Oct 16, 1997 at 07:24:48AM %2B0100 References: <19971015171047.05325@ct.picker.com> <199710160624.HAA10739@labinfo.iet.unipi.it>
next in thread | previous in thread | raw e-mail | index | archive | help
|well you missed the last sentence :) Well, no truthfully I did see that last phrase the first time. :-) It brought back memories of my "enjoyable" strolls through the current driver code figuring out the mods to port the Linux AWE driver into it (and after that, memories of how much time I spent getting the one module committed). That was just porting; we're now talking about redesigning, not just the driver, but the entire working synth subsystem. Time, Talent, and Enthusiasm allowing, I'm with you. I prefer clear code with good comments and an interface and design document so its easy to maintain and extend, particularly when it's done that way the first time. However, if across the group we're lacking T T or E for a rewrite, unless you know of some prohibitive bugs in the existing synth code, it might be a good idea to lean toward working it so that the synth system's existing external APIs can be "plugged in" so-to-speak across a DMZ (your clean code on one side; ugly-but-functional code on the other) into a compatibility API in your rewritten modules. We can leave a "FIXME: rewrite me someday" comment in the synth modules for when TT&E all come along to rewrite them. I guess while you're writing your documentation over the next few days (which will of course be a big help to whoever does the rewrite), you might jot your ideas on the synth redesign in an Appendix. With your knowledge, this should give the person(s) that take this on a big head start. | really I hoped that people |like you could help in adding support for the synth since you |certainly have quite some expertise and maybe even interest in |these features. But of course you also might have limited time and |cannot work on everything. Well thanks, but I actually don't know much about the current sound driver innards. I just did the port of the AWE module, poking around in that current driver enough to accomplish that. In terms of TT&E, over an extended period I could probably find the time. Talent, read "innate knowledge of the old synth systems' inner workings", is severely lacking now but could be developed (with listings, a highlighter, much coffee, and some aspirin :-). However, in rank on the rest of my "things I could be doing with my free time" list, the prospect of redesigning this working subsystem isn't way up there. Someone else more into soundcard coding might be interested in this project though. Randall Randall Hopper: >Luigi Rizzo: >||>|I am wondering if there is anyone who really needs /dev/midi and >||>|/dev/synth, and used them on the old sound driver. If you do, (Uncle >||>|Sam's finger pointed at you) speak up, telling which apps you use >||>|which do not work with the new audio driver, where I can find them, >||>|and how are you going to help in implementing these devices. >||> >||> Does this include /dev/sequencer too? I use this device all the time via >|| >||yes it does. >||in fact I believe timidity is a possible alternative to this. it takes >| >|> I use timidity on my Sparc at work and have used it on Linux boxes. Nice >|> program if you don't have a card with wavetable support (which BTW are dirt >|> cheap nowadays). However it sucks serious CPU, and the sound quality with >| >|i forgot... also CPUs are dirt cheap these days, and last time I >|checked a Pentium133 (or 166 ?) was cheaper than an AWE32... >| >|my point is, perhaps you can invest the money for the wavewable card >|into a much faster (or secondary) CPU. > >So do you mean to say that since there's a program that'll do >better-than-nothing MIDI synth on the CPU, we should rip out support for >the synth devices and for soundcards that support synth in hardware?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19971016194418.46875>