[Table of Contents]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [AV Media Matters] The esoteric edge/ Towards a mediaLESS archive

>Any comments from those of you who might have looked at this model?

In August of 1998 I wrote a posting to the AMIA listserve that discussed
this issue at some length.  It was quite contraversial at the time and
got quite a few comments.  Perhaps it will again.... here it is -

I have some thoughts that are looking towards... A Virtual Archival
Medium -

OK so this is not going to happen in the next couple of years, but here
are some of my futurist type thoughts....

I have been thinking about exactly where all of this information can be
held in the safest possible place, in the safest possible way, and in a
way that it can be readily accessed and used (because information that
cannot readily be accessed will eventually not exist), and I have
decided that no one or two locations (or even three or four) could ever
be really safe and provide access for future eons no matter what the
nature of the storage media was or the technologies were that interfaced
with the media.... So I think that the perfect Archival Medium would
have everything everywhere at once for everyone and not physically exist
anywhere.  This assumes many things.

It assumes that there is almost infinite bandwidth, meaning that on a
practical basis, you can get as much information as you need as quickly
from the sources that hold them whether they be remote sources or local
sources as you can if they were in your facility.  I.e.,. No waiting for
anything.  This is actually not as far fetched as it sounds.  Fiber
optics already have a tremendous amount of bandwidth that is not even
used, the problem is in getting that bandwidth into your home or office
- and this is a relatively short term problem.  Obviously the speed of
the bus and computers are very important as well, but as a practical
matter, once you have reached the level of speed that is required - fast
enough is fast enough... Meaning that the throughput can be defined, and
once it is there it will be good enough for our purposes.  So if a
remote storage site can get you the information as quickly as your local
storage - what does that mean?  It really means that there is no real
incentive to keep information locally, other than factors of cost of
storage - which brings me to....

The assumption that the incremental cost of storage is essentially
nothing - or extremely small.  Think of storage as a large commodity
exchange network that everyone is a member of.  I go out and purchase a
few hundred terabytes, and essentially stick out a sign on the network
saying that I will store your stuff at x cents per terabyte.  Your
computer looks at the world storage market and finds a great deal at my
new storage facility, and you move SOME of your data instantly to my
site.  The SOME is an interesting part, because just like with RAID disk
arrays, you only want enough at any given site so that you could
reconstruct the data if that information was lost or corrupted at that
site (or maybe you don't pay your storage bill on time!)... Similar to a
RAID system whereby you can loose a significant portion of the data and
reconstruct it with the existing data (this is existing proven
technology).  So your computer sits on the "net" and looks at the
storage commodity exchange news and keeps tabs on where the best deals
are.  There is no real difference where you keep the data, because
owning storage yourself does not really mean anything since virtually
everyone is in the storage commodity exchange.  Your system keeps track
of where everything is at any given time - and of course that
information is kept everywhere at the same time as well.  Files the way
we think of them become a meaningless concept.  What really matters is
the STRUCTURE of the information, not the container that it is held in.

Of course the size of the storage is as small as necessary.  Imagine
holding all of the data that would represent (uncompressed of course)
all of the feature films made to date in a container the size of a paper
weight.  So if you actually need to take some information some where
that you do not really have adequate access (highly improbable)- space
and weight are not an issue.

And what about access?  I can give everyone or anyone or no one access
to my information.  I can offer access to my information, and if you
want it, I can charge you for it, similarly if I want your information I
will be charged for it.  So in addition to a storage commodity exchange
there is also an information commodity exchange.  And the fun part are
the finding aids.  Essentially information 'bots that keep going around
the virtual net keeping track of who has what where, and they help find
what you need... in some sense similar to the primitive search engines
we have now, these 'bots can find images and sound as easily as text.
All information is searchable with multiple boolean type tools that are
intuitive and have enough flexibility to allow you to define your search
in terms that make the most sense to the searcher.

And of course the information is safe, because the only real way to keep
information safe is to keep it everywhere...

Sound crazy?  Science Fiction? Well just remember I am the guy that has
these 2" Quad machines that weigh almost 1000 pounds, draw 60 AMPS, need
tons of air conditioning and they were in service only 20 years ago


[Subject index] [Index for current month] [Table of Contents]