FW: DB app recommendations?

["Goldstein, Dan" <DGoldstein -at- DeusTech -dot- com>]

We have Access, but as I said, its limits seem to rule it out. And quantity
of data doesn't seem to justify Oracle.

The writing side of the shop uses Word exclusively -- but as you pointed
out, it would be preferable if the DB speaks XML or RTF, which would make
the data useful for any number of apps.

OCR limitations are well known, but the hard copy's gotta be ingested
somehow.

Editing is the least of our troubles. We have an excellent review system in
place -- we just need a better way to mine the existing, available text
while building a first draft. This is *not* a single-source documentation
project, but rather an attempt to add tools to the existing mining process.

Thanks,

Dan Goldstein

-----Original Message-----
From: Hart, Geoff 
Sent: Tuesday, September 16, 2003 11:01 AM
To: TECHWR-L
Subject: DB app recommendations?

Dan Goldstein wonders: <<My company wants to take a large amount of
proposal-related text (~10K pages), break it up into blocks of between 1
paragraph and 15 pages, and store the whole thing in a heavily
cross-referenced database, so text can be
pulled for use in future proposals.>>

This suggests you may need an industrial-strength database (e.g., Oracle,
DB2) rather than a standard PC system (e.g., Access), but I'll leave
specific recommendations to those with more experience with databases. If
you do opt for something like Access, confirm that it's capable of handling
such large amounts of text. Also look for features such as "exports tagged
text" (e.g., XML or RTF) so you'll have the minimum amount of reformatting
to do once you pour the text into a word processor or DTP software.

You may want to follow up with details on your actual software (seems to be
primarily Word?) and willingness to change (if there's a large productivity
benefit) so our database experts can warn you of any incompatibilities or
efficiencies they've encountered.

<<Much of the text will have to be scanned and OCR'd into the database, so
current doc formats are not much of an issue. What we have electronically is
principally in Word.>>

Watch out for the OCR. Even if it's 99.9% perfect (a difficult target), that
still means nearly 3 typos per 8.5x11 page. To really do this right, someone
must edit the scanned material carefully. Typos are the least of your
problems, though; you also need to be sure the text is relatively
independent of context so it can be reused; for example, you'll find lots of
page numbers cited and hardwired cross-references that must be dealt with
somehow. You'll also have to figure out which version to choose when (as is
likely the case) you find 200 similar "company information" blurbs (for
example): one of these is most current and will work best for reuse, but
some human will have to make the choice of which ones to keep and which to
discard.

Last but not least, hire a good editor to review the results of this
approach to single-sourcing. It's trivially easy (thus, common) to make
small errors with large consequences when you assemble the final manual. I
once owned a Pontiac station wagon whose owner manual had been
database-published in this manner, and several sections were clearly taken
from the manuals for an entirely different type of vehicle--and were thus
misleading or outright wrong for my car. The person doing this editing must
be clearly instructed to worry about more than grammer and wording: they
must first and foremost make sure that even multiply approved and copyedited
paragraphs make sense in the context of the current document. Often, they
won't without human intervention.

--Geoff Hart, geoff-h -at- mtl -dot- feric -dot- ca

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

NEED TO PUBLISH YOUR FRAMEMAKER CONTENT ONLINE?
?Mustang? (code name) is a NEW online publishing tool for FrameMaker that
lets you easily single-source content to Web, intranets, and online Help.
The interface is designed for FrameMaker users, so there is little or no
learning curve and no macro language required! See a live demo that
will take your breath away: http://www.ehelp.com/techwr-l3

---
You are currently subscribed to techwr-l as:
archive -at- raycomm -dot- com
To unsubscribe send a blank email to leave-techwr-l-obscured -at- lists -dot- raycomm -dot- com
Send administrative questions to ejray -at- raycomm -dot- com -dot-  Visit
http://www.raycomm.com/techwhirl/ for more resources and info.