Re: Using FrameMaker to Mothball

[Dick Margulis <ampersandvirgule -at- WORLDNET -dot- ATT -dot- NET>]

Bruce, et al.,

I am in the process of shutting down an entire documentation system for
an engineering project that has been successfully turned over to
manufacturing. (Beginning to look like Brigadoon around the office.)

This is in an FDA-regulated industry, and the integrity of the
documentation is critical. Any given document must be retrievable on 24
hours' notice for the life of the product plus three years. And the life
of the product means when the last US customer finally retires the
machine and the FDA sends a letter certifying that there are no more out
there in use in this country.

Meanwhile, the product support organization will be cannibalizing our
documents to use as the bases for new documents for future models of the
same instrument.

So here is how we handle the mothballing situation:

Every document has a number. The number maps 1-to-1 to a directory
structure as well as 1-to-1 to a physical filing system (not everything
is electronic, but from this point on I'm only going to describe the
electronic filing).

At the document number level is a directory. Inside the directory there
may be any number of files and subdirectories that are all related to
the document in question. Among these there may be readme files if there
is anything ambiguous or complicated about the contents of the
directory. The readme files are plain .txt files that can be opened from
any platform. (Documents may include Interleaf, Word, Excel, PowerPoint,
various oddball application files generated on a variety of platforms,
... you name it!)

In addition, at the top level of each "drawer" directory ("drawers" are
the major branches of the filing tree) is another ascii text file that
is the index for that drawer. This file lists all of the control
information for the documents in that drawer, such as document number,
revision letter, author, effective date, title, file format, location of
physical file. It also includes any notes needed to aid the reader's
understanding of some of the more obscure objects that are on file (like
a box of test tubes, for example).

Both the readme text files and the index files are all written by yours
truly, in my capacity as the documentation control person for the
project. In other words, contributor X cannot just plop some
disorganized mess into the system and walk away. I have to verify that
it meets basic requirements of intelligibility and organization before
accepting and filing it.

All right, that covers the "official" copies.

Next topic:

All current documents (i.e., latest revision) are, if possible,
converted to PDF files. These PDF files are stored in a separate
hierarchy that mirrors the primary one. At the top level of this
mirrored hierarchy is a single index.pdf file created by combining all
of the individual drawer indexes and tweaking it a bit for usability.
This index, which runs close to 200 pages, is itself searchable with a
simple ctrl-F find dialog.

In addition, I used Acrobat Catalog to create a full-text searchable
index of all of the PDF files stored in the mirrored hierarchy.

Okay, what happens with all of this stuff now?

Both electronic hierarchies plus all of the physical files now go to the
company librarian. She has an established procedure for submitting
documents that requires completing a form which she then enters into her
database. In this instance, she will make a single database entry for
each "drawer." describing its contents in a general way and referring
the researcher to the drawer index for details.

Further, all of the paper will be microfilmed by a vendor, and the
microfilm will be stored in the company library for quick access. The
paper and all other miscellaneous objects will be stored by a vendor
(Iron Mountain, in this instance) who specializes in long-term archival
storage. This material will all be in boxes that are tagged and
identified in the library database, so that any document whose original
number is known can be retrieved quickly.

And more...

Throughout the development of the project, the entire process has been
subject both to internal company audits and, theoretically, at least, to
audits by the company that certifies our ISO 9002 compliance. Eventually
it will also be subject to FDA audit. So we have had every opportunity
to clarify what the requirements are and verify that our process meets
them.


So there you have it, a full-fledged mothballing operation if ever there
was one.

Hope this helps,

Dick Margulis


Bruce Byfield wrote:
>
> I had a call from my sub-contractor at my second job-site today.
>
> She had been asked to print off a manual done by the writer who was at
> the company before me. When she looked in the directory, she found a
> mess. Not only had the previous writer not used a book file, but there
> is strong evidence that at least some of the work was done in Word, then
> pasted or converted to FrameMaker. There was only a very confusing
> Readme file to help her make sense of the structure - and she had to
> print ASAP.
>
> Like most contractors, I've mothballed several projects. I've always
> tried to leave my work as finished as possible, and to include readme
> files about directory structures and any quirks in the work, so that any
> other writer could quickly pick up where I left out. This behavior seems
> to me the professional and ethical thing to do.
>
> However, discussing this incident got me thinking about the whole
> process. I'd like to ask the lists:
>
> --do list members agree that an orderly mothballing is a professional
> and ethical imperative?
>
> --what steps do list members think that mothballing should involve? What
> do people think should be involved in a complete mothballing process?
>
> --does anyone make any special use of FrameMaker features to aid
> mothballing?
>
> As I write, it strikes me that this is a part of project management that
> is rarely, if ever, talked about among techwriters, or in seminars and
> classes. Yet it is an inevitable part of any contract job (and of any
> full-time one, if you stay there long enough or leave).
>
> I'll summarize any responses in a week or so.
> ---
> Bruce Byfield, Outlaw Communications
> (604) 421-7189 or 687-2133 x. 269
> bbyfield -at- axionet or bruce -at- dataphile-ca -dot- com
> www.outlawcommunications.com
>
> "My younger ghost is laughing,
> He kissed you against the wall,
> I think he looked a lot like me,
> Would you know him now at all?"
> - Oyster Band
>