Questions about cdrtools-2.01a33
30 second summary :
cdrecord and mkisofs from cdrtools-2.01a33 work fine for me, so far.
Some warning messages about the future worry me, though.
I had a compiling problem which may be due to my system.
Thank you, Joerg, for your work and your endurance.
About the disturbing messages :
cdrecord: No write mode specified.
cdrecord: Asuming -tao mode.
cdrecord: Future versions of cdrecord may have different drive dependent defaults.
Will those future default modes stay suitable for single-session backup
writing ? With ISO filesystems as well as with raw archive files ?
On the fly via stdin of cdrecord ?
I got confused by comparing http://www.cdrfaq.org/faq02.html#S2-19
with man cdrecord -dao -sao -tao . (Andy McFadden distinguishes DAO
from SAO, Joerg does not. Since -sao needs to know the data size
in advance - is it suitable for stdin at all ?)
May i ask for the reason why this change is planned ?
My problem is that my software uses local installations of cdrtools
or even cdrecord-1.X . I faced my little collection with option -tao :
cdrtools-1.10/cdrecord/OBJ/i686-linux-cc/cdrecord: Bad Option: -tao.
cdrtools-2.0/cdrecord/OBJ/i686-linux-cc/cdrecord: Bad Option: -tao.
cdrtools-2.00.3/cdrecord/OBJ/i686-linux-cc/cdrecord: Bad Option: -tao.
cdrtools-2.01a19/cdrecord/OBJ/i686-linux-cc/cdrecord: Bad Option: -tao.
(the directory name cdrtools-2.01a19 was chosen by me, not by cdrtools)
Next announcement :
mkisofs: The option '-L' is reserved by POSIX.1-2001.
mkisofs: The option '-L' means 'follow all symbolic links'.
mkisofs: Mkisofs-2.02 will introduce POSIX semantics for '-L'.
mkisofs: Use -allow-leading-dots in future to get old mkisofs behavior.
Must that really be done ? Honestly, i would prefer backward mkisofs
compatibility over POSIX compatibility. mkisofs and cdrecord are distinguished
personalities who can afford to have their own semantics (like dev=0,0,0).
Looks like www.unix.org wants me to register before i may read their specs.
That under the logo "LIVE FREE OR DIE". [insert Bart Simpson shorts quote here]
Ok, -allow-leading-dots seems to be quite traditional. I'll use that in
future ... but what about my old stuff when it meets future mkisofs ?
Will there be any POSIX semantics problems with -l -R -D in the future ?
When will mkisofs-2.02 be released to the general public ?
About the adventure of compiling cdrtools-2.01a33 with SuSE 9.0 :
It was the first time i compiled cdrtools on my new computer.
Sigh.
After ( cd /usr/src/linux ; make cloneconfig && make dep )
compilation by /usr/bin/Gmake worked until it came to this point :
...
make[2]: Entering directory `cdr/cdrtools-2.01/libscg'
==> COMPILING "OBJ/athlon-linux-cc/scsihack.o"
In file included from scsihack.c:127:
scsi-linux-sg.c: In function `sg_settimeout':
scsi-linux-sg.c:1164: error: `HZ' undeclared (first use in this function)
scsi-linux-sg.c:1164: error: (Each undeclared identifier is reported only once
scsi-linux-sg.c:1164: error: for each function it appears in.)
make[2]: *** [OBJ/athlon-linux-cc/scsihack.o] Error 1
...
I bruteforced that by defining HZ 1000 within scsi-linux-sg.c .
(I do not intend to forward this code to anybody. So i hope
it would be ok for this one single installation.)
My usual way to get HZ is #include <linux/param.h> but i am
not eager to explore any unexpected effects within libscg.
In /usr/include/asm/param.h i seem to have two HZ values to choose from :
#ifdef __KERNEL__
# define HZ 1000 /* Internal kernel timer frequency */
# define USER_HZ 100 /* .. some user interfaces are in "ticks" */
# define CLOCKS_PER_SEC (USER_HZ) /* like times() */
#endif
#ifndef HZ
#define HZ 100
#endif
I would appreciate a short hint if HZ 100 had been preferable in this
situation.
If i understand the usage of HZ in sg_settimeout() then my timeouts
might be too long in this case. Probably better than too short ... i hope.
$ uname -a
Linux ts4 2.4.21-215-athlon #1 Tue Apr 27 00:53:38 UTC 2004 i686 athlon i386 GNU/Linux
After a week of usage i found no runtime problems with 4x and 12x
CD-RW media. CD content verifies nicely with an MD5 check.
Have a nice day :)
Thomas
Reply to: