Re: Multisession DVD creation problem.

["Thomas Schmitt" <scdbackup gmx net>]

Hi,

i wrote:
> >   Media nwa    : 2284912s

Joerg Schilling wrote:
> Note that cdrecord lists this as "tracks" and it would only be useful if all
> drives will return the expected sector number for the read session offset 
> "command". Could you confirm this?

It's in the specs. SPC-3 and MMC-6 are current. I learned from MMC-1,
MMC-5 and dvd+rw-tools (growisofs et.al.). I recorded my findings in

  https://dev.lovelyhq.com/libburnia/libburn/raw/master/doc/cookbook.txt

where in chapter "DVD+R[/DL] Cookbook" i wrote:

  Next Writeable Address is fetched from the reply of 52h READ TRACK
  INFORMATION with track number FFh. (mmc5r03c.pdf 6.27)

(That's the same gesture i use for CD media.)


But i assume cdrecord -media-info takes it from this transaction on the
"Incomplete fragment" which appears as logical track 51 of my old DVD+R:

  Executing 'read track info' command on Bus 12 Target 0, Lun 0 timeout 240s
  CDB:  52 01 00 00 00 33 00 00 28 00
  cmd finished after 0.000s timeout 240s
  Got 40 (0x28), expecting 40 (0x28) bytes of data.
  Received Data:  00 26 33 33 00 07 61 01 00 22 DD 70 00 22 DD 70 00 00 27 D0 00 00 00 10 00 00 27 D0 00 00 
00 00 00 00 00 00 00 04 0E D0

yielding:

     51    51 Blank  2284912    2295103    10192
  ...
  Next writable address:              2284912

The address 2284912 is in the "Received Data" as bytes 00 22 DD 70.

Be aware that not the first blank track marks the NWA but the last
blank track. DVD+R and BD-R permit several open tracks with an NWA each.
All but the last track are allocated by RESERVE TRACK.
See for example a DVD from some strange video recorder
  https://askubuntu.com/questions/1105393/retrieving-data-from-old-dvd-with-open-session


> growisofs is a C++ program that has been written as if it was assembler.

One can hardly contradict this statement. But its SCSI doings are clearly
recognizable. Except some bit rot with Linux, its DVD habits are still
quite flawless.


> Aditionally, there is a lot of incorrect code inside that makes is a miracle
> why it could work with some drives

Well, i used it as tutorial and verified all its SCSI gestures with the
specs. Some things i decided to do different. But i found no truely wrong
SCSI transactions. (With BD it looks not that good. That medium type was
still young when Andy retired from development of dvd+rw-tools.)


> it e.g. sends incorrect SCSI "mode select" commands to the drive.

Indeed ? In which situations ? Where in its code ?


Have a nice day :)

Thomas