Re: Long term archiving - which brand of CDROM do you recommend?
"Thomas Schmitt" <[email protected]>
| Newsgroups | gmane.comp.audio.cd-record |
|---|---|
| Message-ID | <[email protected]> |
Hi, Helmut Jarausch: > I have an LG HL-DT-ST DVDRAM GH22NS40 drive. > It looks as if this drive isn't supported by qpxtool. This is well possible and not the fault of QPxTool. The needed info about internal error detection and correction has to be retrieved by non-standard SCSI/MMC commands. Some burner firmware producers implement such commands. Some even seem to publish the specs. Others don't. Worse: Burner manufacturers buy their firmware and core hardware from each other and create a wild mix of drive names and features. > readcd reports that every sector is in error which I cannot believe. If you can read some correct blocks from the medium, e.g. some human readable text, then this assessment can hardly be true. > dvdisaster -s doesn't report any errors. Is that the same as http://dvdisaster.net/en/howtos10.html ? This looks like a checkread for success with the drive-media internal checksums. I.e. whether the drive indicates read error and how long it needs to deliver the requested blocks. Long read time indicates difficulties and multiple read attempts. > To sum up, I'm confused. That's why i propose to have own checksums apart from the drive-media checksums. My backup tool scdbackup records one MD5 for the whole ISO filesystem. My burn backend xorriso records with -for_backup three MD5 for the ISO session and one MD5 for each data file. The first session MD5 guards the superblock, the second guards the directory tree, and the third one guards the rest with all data file content. dvdisaster seems to offer similar capabilities, with more emphasis on error correction. (I would feel better with two independent volume copies rather than with two redundancy augmented single copies of half the payload size. But that's personal taste.) In any case it should be able to tell you if your medium went bad. Have a nice day :) Thomas -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: https://lists.debian.org/[email protected]