RE: User Manual for a Structured Application

[Dan Emory <danemory@xxxxxxxxxxxx>]

Arnis Gubins asked:

>I was also wondering if you could expand (on the lists for the benefit of 
>all of us) on the pros/cons of this XML approach to database publishing 
>vs. using third-party tools such as DLFormatter, Miramo, Patternstrean and 
>UniMerge. Do you see these applications slowly going the way of the dodo 
>with XML databases?
I would say that producing tagged XML as the output from a database has, as 
it's main potential advantage, Unicode, because of its capability to 
represent all upper-ASCII characters (as well as glyphs that do not have 
upper-ASCII code points, and thus cannot be represented within the ISO 
Latin 1 character set). This advantage, however, is vitiated by the fact 
that FrameMaker 7 is no more Unicode-aware than earlier FrameMaker 
versions, thus any Unicode character that cannot be matched up with a code 
point in the FrameMaker character set will not reproduce on 
import.Moreover, the FrameMaker character set has about 6 locked code 
points used primarily in Central European and Cyrillic languages, and, if 
the XML instances contain Unicode code points for those characters, they 
won't be reproduced either (there are, however, special fonts which can 
resolve those problems with locked code points, but it's doubtful whether 
the Unicode-to-FrameMaker Character set mappings provided in the FrameMaker 
7 product will work with those special fonts). So, the huge advantage of 
Unicode will continue to be devalued until a version of FrameMaker is 
produced which can fully utilize Unicode fonts. Consequently, the main 
advantage of XML over SGML (for the time-being, at least) is that XML does 
not require the replacement of upper-ASCII characters in the ISO Latin 1 
character set with entity references. This is not much of an advantage, 
however, since it would be relatively easy, when outputting tagged SGML 
data from a database, to map in the conversion of upper-ASCII characters in 
the database to their corresponding SGML entity references.

In reality, therefore, it's just as easy to output tagged ISO LATIN 1 SGML 
from a database as it is to produce tagged, Unicoded XML And, if the past 
is any indication, FrameMaker 7 is likely to have bugs in its new XML 
import capability, which is why the application documented by the User 
Manual I'm offering to all comers provides a backup capability to import 
SGML rather than XML.

Now, to address your main question, which is whether the capability to 
produce and import into FrameMaker 7 tagged XML or SGML from a database is 
likely to cause third-party database publishing products such as 
DLFormatter, Miramo, PatternStream, and UniMerge to become obsolete. The 
answer is No. The reason why I've reached that conclusion is that database 
publishing using tagged XML or SGML as the input only works for really dumb 
applications where there is no merge-time requirement to test, manipulate, 
rearrange, and format data fields and data records based on their content, 
and there is no merge-time requirement to add things such as data-derived 
index markers, header/footer markers, headings, and static text to the data 
stream.

In the 8+ years in which I have been doing database publishing with 
FrameMaker and UniMerge, I cannot recall a single instance in which those 
kinds of merge-time operations were not required. In the case of the 
application documented in the User Manual I've offered to all comers, it 
was still necessary, post-import, to manually insert running header/footer 
markers and index markers. When the client approached me about developing 
that XML/SGML import application, I strongly recommended that the same 
thing could be accomplished better and more easily by developing a UniMerge 
application to process untagged data records extracted from the database. 
But the client insisted he wanted to process tagged XML or SGML, so that's 
the application I developed. After completing the application development, 
I'm even more convinced that a UniMerge application would have been a 
better (and cheaper) solution.

Nevertheless, I think that exporting tagged XML from databases to 
accomplish database publishing will become the new hot idea, and if that's 
what clients want, that's what I'll give them.


====================
| Nullius in Verba |
====================
Dan Emory, Dan Emory & Associates
FrameMaker/FrameMaker+SGML Document Design & Database Publishing
Voice/Fax: 949-722-8971 E-Mail: danemory@primenet.com
177 Riverside Ave., STE F, #1151, Newport Beach, CA 92663
---Subscribe to the "Free Framers" list by sending a message to
majordomo@omsys.com with "subscribe framers" (no quotes) in the body.



** To unsubscribe, send a message to majordomo@omsys.com **
** with "unsubscribe framers" (no quotes) in the body.   **