Why on earth would anyone who cares about data preservation use a closed data format?! I would convert everything to plain text, one way or another. If necessary, I would include a text file describing everything (unobvious or not; I would prefer to describe unnecessarily than omit something I thought obvious only for it to be unobvious to the future readers). I don't know which natural languages will still be extant a thousand years on, but, given the number of people today who can understand texts from a millennium (or more!) ago, I would expect such expertise to be available, though probably not common. This could perhaps be improved by including a deliberate Rosetta stone - translate the text into every sufficiently major language live at the time (which today might mean English, Arabic, Mandarin, Russian, and Hindi, to name the ones that come to my mind immediately). It might also help to treat the future readers as first-contact aliens (because, mentally, they very well might amount to that) and include a prelude along those lines; people have given a good deal of thought to that question and their recommendations might be worth reading. As for character-set encoding, I see no alternative but to assume that the future readers are at least as competent at breaking simple substitution ciphers as we are, which pretty much amounts to ignoring the issue. If you're including a first-contact style prelude, it may be able to help. Were the first DVDs burned? I've always assumed they were pressed, more or less like CDs. Am I wrong? Yeah. Compare today to 1913.... /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

When I was fairly new at the CD-R thing, I began purchasing what was then known as Mitsui Gold media, which claimed a lifetime of 300 years. I can still read what I wrote then just fine--but so can I also read the stuff written on bargain-basement "blue" CD-Rs. I got an 8" floppy in yesterday from a machine shop operation who needed their software rescued. This was Maxell "R-V" floppies; something like "Reliable and Verified". The odd thing is that the data was recoverable just fine (written sometime during the 70s), but the floppy *jacket* was disintegrating--it had developed large cracks and was starting to crumble away. This makes me wonder how many years a "1000 year" CD-R will have on it before the substrate disintegrates, rendering the medium unreadable... --Chuck

----- Original Message ----- From: "Jerome H. Fine" <jhfinedp3k at compsys.to> To: "General Discussion: On-Topic and Off-Topic Posts" <cctalk at classiccmp.org> Sent: Saturday, March 23, 2013 7:21 PM Subject: Re: OT: 1000 year DVD storage I burned CDs back when the burners were $1200 and the media was $7+ a piece in bulk. The quality of blanks seems to have gone down quite a bit (faster then price). I still have 15+ year old CDR that work fine, but I don't think new media will last that long (kind of miss the blue dye blanks). DVD recordable media seems to leave a lot to be desired in quality right out of the gate. Don't have a blue-ray burner, how are those?

On 03/28/2013 10:30 AM, stefan skoglund(agj) wrote: I guess I should have said "How old are CD-R's at this point?" Since commercial mass-produced CD's are produced by a different manufacturing process, they probably don't count. --Chuck

On Mar 22, 2013, at 10:49 AM, Mouse <mouse at rodents-montreal.org> wrote: Because sometimes they are prohibited from doing otherwise by their employers (governments are especially guilty of this, since their staffs are often either incompetent or consultants who have a financial interest in prolonging the agony). Often enough, especially with governments, the person put in charge of data preservation is not someone who has the foggiest idea of how to actually go about it. They then have to rely on other people to do it, which is great if you know the right people and disastrous if you only THINK you know the right people. I assumed the question to specifically mean early DVD-Rs, but that could be errant. - Dave

On Mar 22, 2013, at 12:25 PM, Mouse <mouse at rodents-montreal.org> wrote: I think we're on the same page here. I suppose it would be fair to say that the folks who would use closed data formats to preserve data do not actually care about data preservation, so I didn't quite answer the question you asked. :-) - Dave

Burned or pressed, the DVD drive seems to read the DVD equally well. What about the physical properties of the media? I tend to assume that the media of a CD is different from the media for a CD. Does anyone have the physical specifications? Since a DVD drive can read both a CD and a DVD (as well as, for many drives, but not all) burn both a CD and a DVD, the media can't be that different. However, since my DVD drive is able to distinguish between a DVD blank and a CD blank, I assume that the media must be somewhat different. The file structure is probably the most important aspect. Do a DVD and a CD have essentially the same file structure? Since the DVD typically holds ten times the storage capacity as a CD, a 32 bit file structure would seem just about correct for the file structure for both the CD and the DVD. At one point, I burned a CD with both a CD file structure starting at CD sector 16 and an RT-11 file structure starting at CD sector zero. Where the RT-11 partitions started (every 16384 sectors of 2048 bytes or 65536 blocks of 512 bytes), I placed spacer files with originally all zeros and filled in the RT-11 file structures after the CD was burned with the CD file structure and converted to an ISO image on the hard drive. It took a couple of attempts to figure out everything, but after converting the first CD image to an ISO file, I used RT-11 to locate all the files from the CD. The first RT-11 file became the file structure of the CD. As for all types of storage, the CD and now the DVD, seems to be about the most stable and unchanging of all storage types. However, I understand that many current PCs these days no longer have DVD drives. Is that the trend? Jerome Fine

On Mar 23, 2013, at 15:31, "Jerome H. Fine" <jhfinedp3k at compsys.to> wrote: For a mass-production DVD or CD, it's just a flat layer of aluminum between two polycarbonate sides. The aluminum takes an impression from the master before the very thin top layer is added. I don't know what the second layer of a dual-layer DVD is. I imagine the Internet does, though. The DVD has a finer pitch in both track width and bit spacing (not unlike a high-density floppy). My assumption is that the data is self-clocking and that the DVD drive will read both just fine to begin with, but that the disc header tells it what kind of disc it is (much like the header tells a CD/DVD burner what kind of media a writable disc is so it can determine the capacity and burn timing strategy). I don't think the laser has the ability to determine different types of media on its own; the receiver generally just sees ones and zeroes (and if you have a drive that wasn't meant to be used with writable discs, or has been intentionally tuned to work poorly with them, like the PlayStation or the original XBox, it doesn't even see those so well). ISO9660 is the standard for CD filesystems, though its limitations often force the use of extensions to get useful behavior out of it (long file names, more hierarchy, etc.). Early Mac CDs were just HFS volumes, but ISO9660 is fairly standard for OS X. I seem to recall that the VMS distribution CDs are just Files-11 volumes. Audio CDs have their own format ("Red Book", named after the standards document itself), which is a simple TOC followed by raw audio tracks (with forward error correction, of course). You can put any filesystem you'd like on a CD/DVD (for example, UDF was a promising favorite for DVDs, since it has fewer limitations than stock ISO9660), you just have to have something that can read it. Further complicating matters is the fact that lots of consumer drives don't behave well with stuff that's perfectly standard, but not common outside of the PC world. For example, lots of VAXen don't like to boot off SCSI CD drives that won't put out 512-byte blocks, but there are plenty of drives like that out there because most common CD formats use 2k or 4k blocks (or 2k plus a bit for the expansion due to the FEC). It's getting there. Just like floppies. Like floppy drives, though, external ones will probably be available for a long time (at least for those of us willing to deal with USB or whatever comes after). - Dave

On Sat, 23 Mar 2013, Mouse wrote: Chiseled stone has clay well beat. It can handle more abuse - ask Moses why 10 instead of 15. But media life won't help without preserved info about content. Look how long it took to decipher hieroglyphics (they got lucky with a multi-language tablet ("Rosetta stone") But how to read Quipos (sp?) is lost. Stonehenge hasn't been figured out. Microsoft doesn't even remember that GW-BASIC stood for "Gee Whiz"! Who remembers why the .EXE file structure used "MZ" as its flag value? If it was, indeed "Mark Zbikowski", then he did NOT succeed in immortality. -- Grumpy Ol' Fred cisin at xenosoft.com

...and all because Cleopatra was the first one of the Ptolemaic dynasty (and almost the last of the dynasty) who could write and speak Egyptian. Before her, the Ptolomeys spoke Macedonian. Ancient Egypt was largely ruled by non-Egyptians. I've long thought it odd that today we take the lives and artifacts of the rulers as being "Egyptian culture". Addressing Fred's question is "who will care what the MZ at the front of an .EXE file will signify in even 100 years?" I remember all sorts of stuff that seemed at the time useful and important. Mostly nobody cares today. Do you really think that we'll have discrete things called "executable files" that are prepared by a process called "compiling" or "assembling"? Will anyone even think that VisiCalc was a major piece of software? Immortality is a fleeting thing. --Chuck