On 08 May 2001 12:00:55 -0700, Rob Pfile wrote: > Bastien Nocera writes: > > On 08 May 2001 11:04:53 -0700, Rob Pfile wrote: > > > > > > David N. Welton <[EMAIL PROTECTED]> writes: > > > > Rob Pfile <[EMAIL PROTECTED]> writes: > > > > > > >> [use ide-scsi to use your cdrom for audio ripping] > > > > > > > Does anyone have a moment to explay exactly why this makes it work? > > > > > > I think the reason is that there is no IDE cdrom driver. there are, > > > however, lots of scsi drivers in the linux kernel, probably inherited > > > from other *nixes (BSD?). SCSI has been around for a long time and is > > > very popular, so that probably explains the wealth of drivers. > > > > I don't know how I should treat somebody with an @apache.org e-mail > > address saying such huge bullshit. > > I dont know how i should treat somebody who is clueless enough to not > understand who wrote the reply. I posted how to get your cdrom working > for audio, [EMAIL PROTECTED] asked the question about why ide-scsi is > required, and i answered his question. notice the word "probably" in > my reply? that indicates that i am taking a guess at an answer.
My bad. Still doesn't explain why you say that there are no IDE cdrom drivers. Maybe not one that works with the TiPB (although I doubt it considering that the ide-scsi emulation finds it), but there are IDE cdrom drivers. > > There are IDE cdrom drivers in the kernel, and there are a lot of SCSI > > drivers mainly because each SCSI adapter/card interfaces in a different > > way with the system. Although SCSI is popular, Linux hasn't inherited > > from any other Unices because it doesn't share any common code with any > > of them (I'm talking about the kernel). > > Fine, thanks for the history lesson. enlighten us as to how to force > linuxppc to recognize the cdrom drive in the tibook and attach an IDE > cdrom driver to it. > > > > > ['eject' doesnt work] > > > > > > i did strace() it but that was before i understood what was going on, > > > so i'll have to try it again. > > > > The output of eject -v would be useful, probably showing a > > misconfiguration, before resorting to using strace. > > okay, i'll try that. > > > > On the endianness front and bug-wise, it must be that there's no > > > uniform way for drivers to report what endianness they return the > > > audio data in. i cant think of any other reason why cdda2wav would > > > attempt to analyze the data. i dont know if cdda2wav or audiocd reader > > > are under active development, but i'll contact the authors. > > > > cdda2wav works fine on PPC. I don't know about "audiocd reader". > > again, if you read my original post carefully, you'll notice that i > said that it worked fine for many discs, and then i encountered one > for which the same cdda2wav and lame setting produced white > noise. then i READ THE SOURCE to cdda2wav and discovered that it TAKES > A GUESS at the endianness of the data coming off of the cdrom. it > guessed wrong 4 out of six tracks for the disc "hi scores" by the band > "boards of canada". try it for yourself, i bet the same thing happens > on an x86 box. You can use the "-C big/little" flag to force cdda2wav to read the audio in one endianess. > seriously, can't we keep it constructive here? sorry i am replying to > this post, but i just can't help myself :( The "eject -v" wasn't useful ? Cheers -- /Bastien Nocera http://hadess.net
