Re: SGML/XML Justification - was replacement of Frame

[Chris Despopoulos <cud -at- arrakis -dot- es>]

How to justify cost of XML vs Non-Structure?  

First you need to analize your current costs.  I can't do
that for you because my costs will always be different than
yours.  If you can boil it down to cost per page, that would
probably be good.  Then you also need to break out the
difference between initial writing and maintenence...  And
you need to identify pages that are written or maintained
needlessly - reusability issues.  But only you can do that
because your environment is your own.  In other words, at
this stage anyway, structure is inherently vertical.  

Next, analize the projected page count one, two or three
years down the road.  Plug it into your cost per page. 
Reusability is an issue here, too.  Does the company want to
publish online?  How much content comes from data files that
might change without affecting any of the prose?  Stuff like
that.  

You might ask whether any prospective customers require
SGML.  

Then you need to at least have a dream of what the structure
system will do and how.  From that you can estimate a new
cost per page.  Is the result an improvement?  Not
necessarily so, but usually as page count and reusability go
up, the improvement increases.  Here is where you make
savings on redundancy, data mining, maintenence, publishing
to different formats, mix-n-match, etc.

From here you should have an idea of what you're willing to
spend on implementing the system.  Start looking for
consultants who are willing to bid.

Structure is not always the most cost-effective solution. 
That's why the current SGML houses are almost entirely
industrial...  Either they need SGML to submit the product,
or the page count and process see added value.  To get the
added value, the page count and process are usually pretty
scary.

Just two euros...                     cud