Version & content management in Unix / Document management - long

Subject: Version & content management in Unix / Document management - long
From: HALL Bill <bill -dot- hall -at- tenix -dot- com>
To: "'Techwr-l (posting)'" <TECHWR-L -at- lists -dot- raycomm -dot- com>
Date: Sun, 6 Feb 2000 14:01:26 +1100

Reni Waegelein (version management in Unix / Framemaker) and Emily Cotlier
document management / MS Word) are both concerned to find tools to provide
version control (and revision history) where multiple authors need to access
the same list of documents.

The suggestions made (use Word's internal versioning for the Word documents
and Parametric Technology's Windchill) are possible solutions to the
immediate problems, but only scratch the surface.

My experience in this area is based on 10 years trying to manage
documentation processes for a medium/large defence systems supplier
(currently building 10 frigates for the Australian and New Zealand Navies)
across the whole life-cycle of the project (prime contract -> subcontracts
-> the whole suite of maintenance documentation). We have made every mistake
in the book, and hopefully have learned from them. Most of our documentation
has been produced in WordPerfect and then MS Word (now Word 98), and we are
now beginning to implement SGML/XML standards in a serious way. I also write
100% of my own production in Word (or ascii email), because those are the
only tools available to 98% of my readers.

The bottom line from my experience is that MS Word should NEVER be used in a
multi-author environment. Versioning will also always be difficult if all
you have are the ordinary network access controls. I have never experimented
with Word's internal versioning, but I distrust any multi-author use of Word
because its inherent style management problems are multiplied by each new
user accessing the document.

Even if you implement a good file level management system (e.g., PC Docs,
Documentum, etc.) that stores documents in their native Word format, you
will still have style management problems.

The best authoring environment is one that manages documents at both the
document level (versions) and content level (elements of content and
associated styles).

The ideal authoring environment uses a structured authoring tool controlled
by SGML/XML Document Type Definitions like FrameMaker+SGML or Adept Editor
which completely separate content authoring from the application and
management of styles. Authors can be given a WYSYG screen environment so
they can see what their output will look like, but have no access to
formatting functions (which actually makes authoring much easier, once
authors accept they are not allowed to control style). Once authors have
adapted to thinking structurally, a 25-50% increase in productivity should
be achievable by moving authors from an uncontrolled Word environment to the
structured environment.

There are a number of Document and Content Management Systems (DCMS)
available that work in both Unix and Windows NT environments. The ones I
have investigated most thoroughly are XyEnterprise's Parlance (one of the
earliest - http://www.xyenterprise.com/), Chrystal's Astoria
(http://www.chrystal.com/products/astoria/astoria.htm, Texcel's IM (acquired
by Interleaf in April last year - now included in BladeRunner, which in turn
was acquired a couple of weeks ago by BroadVision -
http://www.xmlecontent.com/), and the product SIM we are implementing at
Tenix (developed by our home-town RMIT University - http://www.simdb.com;
http://mds.rmit.edu.au/). There are also some lower end systems that also
understand XML/SGML structures (e.g., Poet -
http://www.poet.com/products/cms/cms.html) I know less about. SIM appears to
offer the greatest scalability and flexibility of all the products Tenix
reviewed, with existing users doing concurrent indexing of gigabytes of data
per hour, delivering documentation to tens of thousands of users, serving
nation-wide web sites, supporting eCommerce sites, and managing legislation
as well as coping with the needs of a documentation intensive industry like
Tenix.

A DCMS used in conjunction with a structured authoring environment, not only
provides full version control at the document level, along with workflow
capabilities to control and track document review and signoff, but it
provides a controlled environment for the discovery and safe reuse of
elements of text (i.e., corporate knowledge). In other words, common texts
need only be written (and managed) once, for safe reuse in many deliverable
documents. In a business like Tenix's, where there are high levels of
redundancy across documents, we expect the DCMS system to reduce our overall
authoring requirements by another 25-50% on top of savings to be achieved
from moving from Word into a structured authoring tools.

If you are not prepared to give up MS Word as an editor, BladeRunner can
supposedly impose DTD controls on Word's styles, or RMIT's SIM can parse
Word's RTF format for automatic conversion into XML for storage (as is done
in the legislation management system). Where SIM is concerned, when the
document is checked out again, the Word styles are always automatically
restored to established standard for the document type, so each author
begins with a clean slate. We also used SIM (with some programming in its
integral scripting language) to convert our WordPerfect maintenance routines
into SGML (by first converting the WordPerfect documents into RTF) with 97%
converting automatically from the uncontrolled WordPerfect environment.

If you want more information, check the Techwr-l archives, and search for
"hallb", which should recover my earlier postings on the subject. I will
also attempt to find time to respond to individual requests.

Bill Hall
Documentation Systems Specialist
Naval Projects and Support
Tenix Defence Systems Pty Ltd
Williamstown, Vic. 3016 Australia
hallb -at- tenix -dot- com





Previous by Author: multiple TWs for a project
Next by Author: Re: XML (and chunking)
Previous by Thread: UPDATE - Friendly Faces of TECHWR-L
Next by Thread: SUMMARY - Screen Capture Software


What this post helpful? Share it with friends and colleagues:

Sponsored Ads


Sponsored Ads