[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: cdrecord-2.01.01a21 refuses work on Linux 2.4 if non-ide-scsi DVD-ROM is present



Hi,

> >   http://scdbackup.sourceforge.net/cdrskin-0.2.6.pl01.tar.gz
> It still does not have a working "configure".

If that happens on Linux with a 2.4 or 2.6 kernel
then this is a defeat for our project (and for
autotools, possibly).

I would appreciate a complete log of the failed
configure run. Like from:

  ./configure 2>&1 | tee -i /tmp/cdrskin_configure_log

Helpful would be the output of your system's

  uname -a

plus the MD5 of the tarball which you unpacked.


> Well, Linux is not the world

Indeed it isn't. But it happens to be the operating system
of those whom i know and who do not use Microsoft Windows.

libburn's lowlevel SCSI adapter is prepared for porting now,
but we would need system experts, test systems and testers.
(A FreeBSD port sketch for CAM exists. The task for the system
expert would be to introduce us to the equivalent of Linux sg
or FreeBSD CAM.)


> It looks like you are misunderstanding the libscg interface.
> Have a look at the syntax definitions in the source files....

I will make use of your invitation. But for now i did
not try to understand your sources but only studied the
use cases of cdrecord and its recognizable behavior in
particular situations.

>From the user point of view, cdrecord's approach towards
drives was much superior to libburn's (see ill DVD-ROM).
libburn's builtin ability to address a default drive
did not outweight the disadvantages and risks of a mandatory
full bus scan.
This ability is convenient and cdrskin makes use of it
- but only if there is no drive given explicitely.


> > ill DVD-ROM which stalled libburn's
> > also stalled  cdrecord dev=ATA -scanbus on a Linux 2.6 kernel.
> This looks like a Linux bug.

Stalling is not a nice behavior on open(), indeed.

But see how elegantly cdrecord avoided any trouble
by just not touching the ill drive. 
Ok, dev=ATA -scanbus was broken. But no reasonable user
would run that command on a system with ide-scsi burners.
And even if a normal -scanbus would have included a
dev=ATA scan and would thus have failed: one still
could have used both drives dev=0,0,0 and dev=0,1,0.

The new approach makes cdrecord more prone to external
disturbance. Especially with those use cases where no
option -scanbus is present.


> "make relink"

That one is new to me.

> read the documentation

I already got square eyes from acroreading SPC/SBC/MMC
specs in various draft versions ...
Allow me to stay with try-and-error as long as i succeed
in less than three attempts :))


Have a nice day :)

Thomas



Reply to: