[time-nuts] Archiving Data

Chris Albertson albertson.chris at gmail.com
Sat Jan 8 03:31:29 UTC 2011

You are 100% correct that no digital media is "archival" and all of it
will fail or there will be no machine that can read it.  Certainly
everything we have will be junk in 100 years.   But digital data is
not the same as digital media.  Data can be perfectly copied.  We can
make many copies.   For example at work we have a disk array.  This
device spreads data over many drives in such a way that if several
disk drives fail no data is lost.   But we have three of these arrays
in three different cities and they are kept in sync with each other
usig high speed data lines.  So if there were two natural disasters
and both Denver and Los Angeles were wiped out we'd still have one
disk array.   On top of this they make periodic copys of the array to
tape and send the tape copys to different locations.   The tapes get
re-used and don't have to last forever

Before long EVERYONE'S data will be on devices like these.  The trend
is to move data to the "cloud".  It really is better because then my
data is accessible from any computer.   We are moving to hand held
computers at a very fast rate.  These machines are connected t the
'net 24x7.

I can't afford three disk arrays for my personal data but I to have a
system where I make incremental backups (never over writing old data)
and I rotate my disk drive through an off-site location.  I retire
disk drives after a few years and replace them with newer larger
drives.  I really do not care to much how long media lasts.  I figure
it has a 3 to 5 year lifespan.  Mostly I retire the media before it
fails although I did have a disk fail last year while in use.

Yes there will be a short historic period from the late 80's to maybe
the end of this decade where much data was "managed" by technically
illiterate consumers in there own desktop computers.  But those days
are numberedand soon most data will be in the "cloud" where it is
replicated to the point where and atomic war would not cause it to be

Technically all that has to happen to keep data forever is that the
rate of replication is greater than the rate of media failure.

 I just saw an ad today for a 2TB disk for $99.  Buy a few of these
every years and make a goal to replicate your data twice was fast as
you expect the media to fail.

Please do not argue that one type of media lasts longer then some
other.  If your plan requires long lasting physical media if will
fail.  No media lasts long in a fire or major earthquake or to
vandalism.   The top reasons for lost data is actually theft of the
equipment and human error.  "archival" media does not even address the
top reasons for loss

On Fri, Jan 7, 2011 at 12:59 PM, Perry Sandeen <sandeenpa at yahoo.com> wrote:
> List,
> I apologize in advance for my long posting
> Several weeks ago I posted what were my attempts to save data and my school-of hard-knocks learning curve.  Unfortunately several posters just had to nit-pick the process I had used and started a long series of posts and counter-posts about the process while totally missing the message.  So I’m going to walk through this again hopefully for the edification of the majority.
> Just several days ago a 10 year old Canadian girl discovered a super-nova while studying photographic images.  Observations by the world’s two most powerful earth based telescopes confirmed here discovery.
> Now consider the case of the Antikythera mechanism.  It was close to 2,000 years ahead of what was eventually developed in Europe.  Most likely we never knew about it was that the library in Alexandria Egypt was joyfully burned three times by religious idiots [see Wikipedia].
> This mechanism was so complicated and accurate that as least passing knowledge, if not some or all of its drawings, would have been there.  In context to the science of that time, it ranked up with what the Hubble telescope accomplished for science today.
> Which brings us here to today.
> Governments and private businesses are storing millions of tons of written documents in the evacuated chambers of salt mines of the world [see Wikipedia].  This does beg the question of where the inventory lists are stored.  The reason for this is that no other archival grade of mass storage really exists.  The last method that I’m aware of is black & white polyester microfilm which is rated at 500 years.  With the almost total transition to digital cameras, there is no financial incentive to produce the necessary film stock to continue that process.
> NASA has lost large amounts of acquired data as no equipment now exists to read the information.
> Current CD/ DVD media is no solution.  A 2004 report published in the Journal of Research of the National Institute of Standards and Technology entitled "Stability Comparison of Recordable Optical Discs-A Study of Error Rates in Harsh Conditions."  You can read it yourself at http://nvl-p.nist.gov/pub/nistpubs/jres/109/5/j95sla.pdf.
> Until someone invents a stabilized glass DVD and perhaps a holographic laser beam to create the needed pits W/O chemicals (embedded gold maybe?) archival data storage W/O paper is a crap-shoot.
> So where am I going with all this?  Glad you asked.
> No one knows if their or others data will lead to a new discovery or process.  Da Vinci certainly didn’t.  It can however; lead to learning that can then can be taken to the next level of knowledge and invention.
> Words cannot express my gratefulness to those who have taken the considerable effort and expense to post science information and support technical lists such as this on the net as well as the posters who have kindly shared their knowledge with us.
> I just hope it doesn’t get lost.
> End of Rant.  I now get off my soapbox and return you to your normal programming.
> Regards,
> Perrier
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.

Chris Albertson
Redondo Beach, California

More information about the time-nuts mailing list