[Framers] Using Frame as a little CMS

[David Artman]

One idea for you that I hope will be useful for these issues (works for
us!):

- There is no way, other than our spreadsheet, to see where the text
insets are used. So if inset A is used in book A, B, and C, we can't
tell that other than through the spreadsheet. You can't click the text
inset and see where it's used.

First, you do need a book-of-books with all of your library in it.
Usually not too daunting; and if your library is vast, make several
divided by whatever scheme you like.

Then, simply append a "file-killer' suffix to the inset that you are
testing for locations (e.g., rename "inset_How-to-Foo.fm" to
"inset_How-to-Foo.fm.HIDE").

Finally, do an Update on the book(s)-of-books and see in the log where
you get errors because of the now-'missing' inset file. [I know that's a
bit tedious for EVERY inset file update, but if you REALLY want to be
sure of where you're hitting existing books....]

- Also, if we need to find information and we search the book or books
for that string, we can't find it as text insets aren't searchable. 

Easy as pie! Make a 'warehouse book' which contains every single FM file
in your library, solely for the purposes of doing global searches.
Again, if your library is vast, make a set of them divided by whatever
scheme you find most useful. In fact, a categorization scheme could also
help you with inset discovery (i.e., writers checking whether something
they want to add to a publication is already written and reusable): The
easier it is to narrow down a search prior to turning FrameMaker's Find
dialog loose, the better for efficiency.

Hope this reaffirms your strategy! And don't take any duff from folks
who think managing hundreds of insets is a bad idea--I've seen libraries
with thousands....
David

P.S. You missed a REAL and intractable issue: I've found sometimes that
cross-references just don't like text insets; and so I often advise a
strategy where headings are not the 'top leaf' of inset content, and the
actual-text-in-files must be targeted for cross-referencing (i.e.,
heading must be typed every time, then inset inserted). As such,
text-inset content can only really internally cross-reference (i.e.,
inside its own inset) or cross-reference between 'real' text in the
files that make the actual deliverable. Not a giant issue, but
non-optimal compared, say, to DITA conreffing.
DCA