David Rothbard
David.Rothbard@ipst.edu

From my reading it seems that magnetic media has an archival lifespan of about 10 years whereas data archived on CD-ROM has an archival lifespan of 30 years. Since I plan on retiring in 25 years, my bet is to archive on CD-ROM and not have to repeat the archival process every 8-10 years.

The upside of magnetic media is the speed at which you can archive I guess, the latest CD-R drives appear to need about 40minutes to "burn" the data on disk.

Kevin McCarthy
KJMCCARTHY@bmg.bhs.uab.edu

I agree that magnetic media has a lifetime of 10 years or so, whereas optical media is far longer, say 30 years. However, I don't believe I have *ever* accessed original data (say negatives) for any useful purpose even 5 years after original recording. Magnetic media has an overwhelming advantage in daily usage, because of its speed, portability (the frisbee test), and competitive price of media and drives. However, for the convenience of our users, in our lab we record archival data (TEM images etc) at present on CD-ROMs, and provide either Zips or CDs to our users to carry their data home. But I certainly do not worry about how I will read that archived data 25-30 years from now, because there is no question that many generations of drive devices will supplant anything that is currently available. We no longer use enlargers to print our negatives, but have scanners that handle transparencies to use if we want to make some new prints. When the time comes, we will transfer our archived images and other data to the next generation of storage devices.

BTW, recently I installed a 400K diskette containing data from May 1984 recorded from my old 128K Macintosh. My Powerbook Duo288c running MacWrite II opened a MacWrite file with no problem, and a recent version of MacPaint opened an old MacPaint file. Of course, I had absolutly *no* interest in any of that data after nearly 12 years.

allardlfjr@ornl.gov
Larry Allard

Pinnacle Micro claims 30 years for its magneto-optical storage media.

John Libert
jlibert@cpcug.org

Dear Larry, list, et. al,

I remember an article in Scientific American about a year ago which mentioned responsible estinates that were MUCH MORE PESSIMISTIC: and on the order of 2 years for tape, 5 years for magnetic disk and about 10 years for CD.

Todays recording densities make for far more fragile records. in the "old days" ten years ago, the energy used to store each bit was 10's of 1000's (!!) times greater than that used today. RAM designers, as a paralell example, are discussing quantum limits ( 1 electron) of data cells.

This is NOT a simplistic topic, but one lying under layers and layers of systems design.

Regards,
Ed Monberg
e-mail: em@mediacity.com

Also, you should read the 'fine print' when it comes to statements about media lifetime estimates. DAT DDS tapes, for example, have a 10 year estimated lifetime. What that actually means is that if you use a tape once to record and put it on a shelf in a temp. and humidity controlled environment, you are likely to be able to read it back in 10 years. Divide that by some factor if stored under less than ideal conditions. I was interested in archival use, rather than backups, where I may need to read back the data many times over that decade. Ask about how many passes of tape over the heads, including rewinding and forwarding, the tape is rated for, and you get a figure that may translate into only a few years of use (depending).

Data format is another problem. I think the ISO CD-ROM standard is in wide enough use to rely upon, but for magneo-optical disks, my experience has been that changing SCSI controller, operating system or other software, even ( with our Pinnacle MO drives ) chaning from one version of DOS to another, is enough to make the data effectively unreadable. ( The documentation necessary to write a program to retrieve that data is available, but not from Pinnacle -- we had to talk to their suppliers to get any useful information. ) I was also told by Pinnacle that they had changed suppliers for their SCSI controller cards, and if we bought a newer model, they would not guarantee that we would be able to read old data on the new drive. I don't know that Pinnacle is any worse that any other supplier -- that's just who I had my unhappy experience with. However, none of the salespeople will bring up the subject, and few will even understand your concerns, if you try to ask them how likely you are to be able to read your data 10 years from now.

ISO 9660 CD format and 'tar' files on DDS tapes (in that order) are the only formats in which I have much long term (decade) faith.

Steven D. Majewski
sdm7g@Virginia.EDU

I'm in on this one with my 2 cents worth.

In our lab we've gone with a CD-writer (i.e. that's what I decided on). The reasons are outlined below;

1) Cost.
Obviously a blank CD (at Aus$14) is far cheaper material cost than film, with all the labour savings as well. I don't recall the costs of Zip or Jaz discs but I have it in my mind that a CD is still far cheaper than these as well.

Hardware costs to my clients is also cheaper in as much as almost all computers that have come into the Australian Museum over the last 12-18 months already have CD drives. This makes CD's a more 'universal' media, in our place at least.

The big cost is the CD-writer but I have no regrets about getting a top-of-the-line model (a Philips CDD 522 - my opinion only - no vested interests). Of course now everyone wants to use it for all sorts of things. I don't mind because, at 600-odd micrographs per CD, the writer sits idle a lot of the time. And if you can get other departments to chip in on the purchase price, well...

Besides, the cost of CD-writers and software has plummetted over the last 2 years, from roughly Aus$12-15K for a package to $5K when I bought mine 6 months ago and probably much less by now. Not REALLY expensive for what it delivers.

2) Archiving.
The general consensus seems to be that CD's will cough up data a lot further down the track than magnetic media and are a bit more tolerant of dust storms, passing magnets, strawberry jam/jelly incidents.

3) Future Viability.
Who can say for sure? My impression is that with the music industry 'locked-in' on CD's that they'll be more stable than most other formats. Sure there will be some changes, like when the blue-laser hi-density video CD's arrive. But I'm hopeful that the new generation readers would read the old stuff too. And once you can fit a movie-length video on a CD, I don't envision a lot of industry pressure to up the volume again for a while.

Summary.
CD's are about the cheapest, most universally readable, hi-volume media going around. Who cares if a client only wants to take 20 micrographs and has a lot of blank disc space? It's still much cheaper than instant film (no names mentioned). There probably is a very good niche for Zip drives for quick transfer of a reasonable fistful of images between computers/departments/work-to-home/etc. But if I had to make the choice again, it would be CD-writer first, maybe Zip/Jaz when I had some more money to spend.

Geoff Avern
geoffa@amsg.austmus.oz.au