Hi! It seems that I'v destroyed a partition with "readcd" The command line was: "readcd -w f=img.raw", i.e. I forgot to specify the "dev=x,y,z" part. The write attempt aborted with an error message about writing to "dev=0,0,0". And on my system this is "/dev/sda". After reboot Linux told me that "/dev/sda" had no valid partition table. At the writing attempt none of the partitions of /dev/sda was mounted. At least I assume it was "readcd" because as far as I remember I did no othing else that could destroy a partition. With Linux fdisk I restored the partition table manually. Before the disk had three partitions (sda1, sda2, sda3). After restauration of the partition table I was able to mount "/dev/sda3". The data seem to be there but I haven't checked the filesystem yet. "/dev/sda2" was an unused SWAP partition which I haven't checked either. But "/dev/sda1" is gone! I wasn't able to mount it. I was using Linux 2.3.26 with cdrecord-1.9 binaries. I assume that with a missing "dev=x,y,z" "cdrecord" would have asumed as well "dev=0,0,0" and would have destroyed the partition. Am I right?????? Now my question: Is there a way to make such acctions error proof against unintended writing? And a question for the developers (i.e. mainly J�rg): Wouldn't it be possible for cdrecord and readcd "to recognize" what kind of device the write target is (disk or cdwriter) and that they abort writing if it is a disk? Friedrich -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
