Dear All,
    I have followed the progress of the discussion called "Gold CDs" right
up to the present day, and one factor still seems to me to have remained
unsaid.
    Whether one attempts to label CDs my means of Wave file-headers (of
which there are several types), or HTML, or even the MPEG7 standard (does
anyone know what happened to *that*?), one always needs *software*. And
there is always copyright in software (in some places, *permanent* copyright
in software!)  And this may force each future user to write his (or her) own
software, to circumvent problems when the copyright software is no longer
available.
    The only way I know to escape from this heffalump trap is to have a
spoken announcement (or announcements) on preservation audio CD-Rs. Then, so
long as the CD can be played as an audio storage medium, the information
about it can also be played. This was a policy I myself introduced to
circumvent all the "solutions" discussed on this thread, and it is still
being done. However, it is also fair to say that some preservation jobs were
so difficult to document that announcements might take longer than the
playing-time of the material! For this reason, the procedure has now evolved
into announcements sufficient to tie the transfer to the original recording,
and to the appropriate "archival" inlay-card in the event that discs and
cases become swopped.
    Nevertheless, having the gift of the gab myself, I should prefer the
whole information to be recorded using speech; but I can quite see that some
operational stuff may not be as loquacious as I.
Peter Copeland
Former Technical Manager, British Library Sound Archive

-----Original Message-----
From: George Brock-Nannestad [mailto:[log in to unmask]]
Sent: 26 September 2004 14:27
To: [log in to unmask]
Subject: Re: [ARSCLIST] Gold CDs

From: Patent Tactics, George Brock-Nannestad

It is remarkable that infinite faith is put into the future ability to
provide readers for the media we store today, but the same faith is not put
into the manufacture of media.

(see Don Cox' response to me, I won't quote it)

I would expect that the level of deep knowledge of those who write about
media and those who write about machines with such assurance in ARSCLIST is
quite inadequate to provide either: none of the contributors to this thread
is either a CD manufacturer (some of the processes are trade secrets, too!)
or the designer of high-speed computer components. However, it only takes a
bicycle repairperson to create a replay machine for Welte, Duo-Art, or
Ampico, and a bicycle repairperson with a magnifier to create a pickup for
mechanical records. It helps to be able to read the technical literature
before starting the projects, however.

I think that anybody who distrusts media but trusts future machines
approaches the field in an ostrich-like manner. It is short-term nice.

Let me instead propose a responsible approach: let us agree on a reader
configuration that we believe will be the one we would like to use in a
distant future to read our long-life data discs of today. Let us then
acquaint ourselves with the components that go into it - oh shit, it is only
these two huge LSI blocks! Well, back to earlier literature; at some stage
we
get back to ordinary ICs (this would be in literature from the 1980s) having
properties that are mentioned in published data sheets of the period.
Fortunately we have the Red Book to guide us in interpreting what all these
components are actually performing. We may hence provide a model circuit
board assembly fitted with components performing known tasks (known to the
individual switch level!), and we can thereby completely specify a replay
machine - for Compact Disc Audio.

Now comes the responsible, long term part: a special committee under ARSC
(and its successors) keeps track of the delivery situation regarding all the
necessary components for this model circuit board assembly, and when they go
out of production, oversees the design of replacement components or
equivalents defined as sub-assemblies of components still available. I do
not
mind if an FPGA solution is used (see below for a definition, which was not
provided by Don Cox).

There should be at least an annual review of the supply situation, and the
fundamental operation of a design would thereby be safeguarded. A special
focus (excuse the pun) should be the optical assembly, sitting in its auto-
focussing mount. From time to time a project (perhaps a student project) of
actually building a replay machine should be tried to see if the design
works
in practice. This work is continuous and should go on for as long as anybody
believes in the data that is provided on our long-life data discs of today.
It may be an official duty for a  national archive to do this, but it may be
farmed out by funding the ARSC committee.

What I have described above was an important activity of a lot of industries
in the industrial age: ensuring that the product could still be
manufactured,
although the supplies changed. I do not remember the English term for the
responsible department, but it was that department that created the most
revisions to the drawings in the drawings store. As an example I might
mention that although Lyrec of Denmark has not manufactured FRIDA (a
fabulous
professional tape recorder) for about 10 years, they still have a system for
maintaining the design, and this means that the moment that a sufficient
number of orders turns up, they would be willing to manufacture, without
having to start from scratch.

Now, if you do not like the thought that you need to refresh, clone, or
migrate your machines just as much as your data, then there is only one way
out: see to it that you have a system that will migrate your data for you,
just the way that large computer backup systems works.

Kind regards, and do be realistic!,

George


-----------------------------
FPGA:
Field Programmable Gate Array.
A specialty integrated circuit which can have its logic and circuit topology
changed in the field (versus the silicon factory) via a number of different
electrical programming methods. The alternative is a standard cell or gate
array design which is manufactured with one circuit topology that cannot be
changed. Devices are not programmed by downloading software to them. Rather
a
manufacturer specific data stream is downloaded into the part to cause
specific, hardwired, logic operations to become interconnected to form the
desired logic and memory operations.
(a quote from www.aspenlogic.com/glossary.htm)


**************************************************************************

Experience the British Library online at www.bl.uk

Help the British Library conserve the world's knowledge. Adopt a Book.
www.bl.uk/adoptabook

*************************************************************************

The information contained in this e-mail is confidential and may be legally
privileged. It is intended for the addressee(s) only. If you are not the
intended recipient, please delete this e-mail and notify the
[log in to unmask] : The contents of this e-mail must not be disclosed or
copied without the sender's consent.

The statements and opinions expressed in this message are those of the
author and do not necessarily reflect those of the British Library. The
British Library does not take any responsibility for the views of the
author.

*************************************************************************