Dominick A. DeFlorio wrote:

> We must move to structure to meet our future XML goal, but are
> unfamiliar with both the transition and method to do so.  We are also
> unfamiliar with the possibilities and cost involved. We are merely
> looking at all of the possibilities and the long term value for our
> dollar.

Here's how I'd do it:

a) Design the structure - if you aren't experienced, don't do it 
yourself and don't buy the line that you can do it with DITA or DocBook. 
Get a professional to do proper analysis and design, including 
documentation about how to use the DTD or schema. This is a critical 
step - don't scrimp here.

b) Save all of your data out of FrameMaker as XML - don't use FrameMaker 
as a migration tool because if your structure evolves based on 
infrequent cases, you'll end up spending too much time trying to 
re-baseline your dataset.

c) Use XSLT to convert from XML to your target structure - if you find 
that you need to make changes, make them and re-run the whole dataset, 
so you can be certain that all documents are consistently handled. Get 
help with this if you're not experienced, otherwise you'll get in a 
mess. Be prepared to also make changes to the data manually - the 
alternative is to loosen the structure in the DTD or schema, but that's 
a last resort. Make the data consistent, and be ruthless about it. While 
you're at it, learn not to be scared to work with native XML - angle 
brackets don't bite.

d) Build the FrameMaker application - concentrate your own efforts on 
the part of the process that you're most familiar with and learn the 
parts that will give you the most benefit.

e) Train your users - give them as little information about XML as you 
can get away with. All they need to know is that there's a mechanism in 
the background that ensures that the documents are structured 
consistently with the rest of the dataset.

Items a) and c) will cost you money, but it's well spent. Do it properly 
from the start and you'll only do it once - try to do it on the cheap 
and you can spend the money next time around. I've seen it more times 
than I could count in the 15 years that I've been involved with SGML and 
XML conversions. Plenty of very bright people have tried to migrate to 
structure on the cheap - so many smart faces, so much egg.


Marcus Carr                      email:  mcarr at
Allette Systems (Australia)      www:
"Everything should be made as simple as possible, but not simpler."
        - Einstein

Reply via email to