Bug#268621: Reproducable in parts
Eduard Bloch <edi@gmx.de> wrote:
> Moin Joerg!
> Joerg Schilling schrieb am Sonntag, den 27. Februar 2005:
>
> > If you like to have a decent CDDA extraction you need to use
> > generic SCSI and this is done by using the SCSI address syntax
> > instead of filnames.
>
> Oh my...
>
> man causality
> man partial_order
It is really bad to see that this poor person still has no clue and
still refuses to read man pages :-(
In order to find out whether this person is not the only one who
has problems to understand simple facts, let me elaborate the
previous mail...
On 2005-02-26 Joerg Schilling <schilling@fokus.fraunhofer.de> wrote:
> dev=ATA and dev=/dev/hdX are definitely not supposed to do the same thing.
Background is here that cdda2wav documents that dev=/dev/hdX
tells cdda2wav to use "cooked audio ioctl's" like:
ioctl(fd, CDROMREADTOCENTRY, ...
instead of sending SCSI commands directly to the drive.
man cdda2wav also tells you that most OS behave really bad
with "cooked audio ioctl's" and for this reason recommends to
enable to use Generic SCSI by using the SCSI address syntax
together with the dev= option (e.g. dev=6,0).
This resulted in a question:
> What is the correct devicename on Linux 2.6, if I want DMA?
And I did reply that I would expect to see DMA for all
ways to send Generic SCSI commands on Linux and that the
problem is that the Linux kernel developers refuse to
fix the bugs that prevent DMA from happen with most
interfaces. I also replied that a decent OS only needs
one single interface to send Generic SCSI commands.
Another questin was:
> dev=ATA is not documented in cdda2wav(1) (from 2.01.01a01).
Well, I was asuming that people know that all SCSI tools
use libscg and in case a single man page does not contain
_all_ information it makes sense to check other man pages
(e.g. the man page from cdrecord).
I also asume that Linux users _know_ what to do....
I am sorry but I am not responsible for the fact that
the Linux kernel developers do not like to evolve an
existing interface that would give thet best result
but rather introduce new and unneeded new interfaces that
include new bugs that are not fixed also......
> You try to force you own design ideas by any means (here: ignoring the
> actual question and presenting your biased definition as a fact) and in
> the same breath you accuse Linux kernel developers of not caring about
> their users. Who t.f. do you expect to listen to you after a such
> action?
Well, in former times it was possible to see desigh ideas in the Linux kernel.
This it no longer true since at least 2 years.
What you see instead is total confusion.
Jörg
--
EMail:joerg@schily.isdn.cs.tu-berlin.de (home) Jörg Schilling D-13353 Berlin
js@cs.tu-berlin.de (uni)
schilling@fokus.fraunhofer.de (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
Reply to: